#  @aditharun_ Adith Arun Adith Arun posts on X about $pasg, $tecx, $rvmd, in the the most. They currently have [-----] followers and [---] posts still getting attention that total [------] engagements in the last [--] hours. ### Engagements: [------] [#](/creator/twitter::1037546970154459136/interactions)  - [--] Week [------] +28% - [--] Month [------] +8.70% - [--] Months [-------] +2,185% ### Mentions: [--] [#](/creator/twitter::1037546970154459136/posts_active)  - [--] Week [--] -26% - [--] Month [--] +2.40% - [--] Months [---] +261% ### Followers: [-----] [#](/creator/twitter::1037546970154459136/followers)  - [--] Week [-----] +1.20% - [--] Month [-----] +3.10% - [--] Months [-----] +355% ### CreatorRank: [---------] [#](/creator/twitter::1037546970154459136/influencer_rank)  ### Social Influence **Social category influence** [stocks](/list/stocks) [finance](/list/finance) [currencies](/list/currencies) [cryptocurrencies](/list/cryptocurrencies) [technology brands](/list/technology-brands) [countries](/list/countries) [social networks](/list/social-networks) **Social topic influence** [$pasg](/topic/$pasg), [$tecx](/topic/$tecx) #5, [$rvmd](/topic/$rvmd), [in the](/topic/in-the), [$prax](/topic/$prax), [$imrx](/topic/$imrx), [$fulc](/topic/$fulc), [$rzlt](/topic/$rzlt), [$lly](/topic/$lly), [has been](/topic/has-been) **Top accounts mentioned or mentioned by** [@plainyogurt21](/creator/undefined) [@3primeanalytica](/creator/undefined) [@houndcl](/creator/undefined) [@avidresearch](/creator/undefined) [@capitalshipyard](/creator/undefined) [@princetongb](/creator/undefined) [@therealryc](/creator/undefined) [@bananaoncology](/creator/undefined) [@adamfeuerstein](/creator/undefined) [@drsamuelbhume](/creator/undefined) [@zhaoweiasu](/creator/undefined) [@financebully](/creator/undefined) [@naturemedicine](/creator/undefined) [@researchpulse1](/creator/undefined) [@medstudentinvst](/creator/undefined) [@rohitsjhawar](/creator/undefined) [@applehelix](/creator/undefined) [@emcclintock](/creator/undefined) [@bermudadude](/creator/undefined) [@biotech2k1](/creator/undefined) **Top assets mentioned** [Passage (PASG)](/topic/$pasg) [Revolution Medicines, Inc. (RVMD)](/topic/$rvmd) [Praxis Precision Medicines, Inc. (PRAX)](/topic/$prax) [Immuneering Corporation (IMRX)](/topic/$imrx) [Eli Lilly and Company (LLY)](/topic/$lly) [AstraZeneca PLC (AZN)](/topic/$azn) [Novo-Nordisk (NVO)](/topic/$nvo) [uniQure N.V. (QURE)](/topic/$qure) [Morgan Stanley (MS)](/topic/morgan-stanley) [BEAM (BEAM)](/topic/$beam) [Prime Medicine, Inc. (PRME)](/topic/$prme) [Elicio Therapeutics, Inc. (ELTX)](/topic/$eltx) [Regeneron Pharmaceuticals Inc (REGN)](/topic/$regn) [Arvinas, Inc (ARVN)](/topic/$arvn) [PMV Pharmaceuticals, Inc. (PMVP)](/topic/$pmvp) [BridgeBio Pharma, Inc. Common Stock (BBIO)](/topic/$bbio) [Alphabet Inc Class A (GOOGL)](/topic/$googl) [Sarepta Therapeutics, Inc. (SRPT)](/topic/$srpt) [Viking Holdings Ltd (VIK)](/topic/$vik) [Xenon Pharmaceuticals Inc (XENE)](/topic/$xene) [Novartis AG (NVS)](/topic/novartis) [Merck & Co., Inc. (MRK)](/topic/$mrk) [Fractyl Health, Inc. (GUTS)](/topic/$guts) ### Top Social Posts Top posts by engagements in the last [--] hours "Does the threshold for low total testosterone need to be re-evaluated New in @Clin_Chem_ADLM @ThomasDurantMD @ClinChemJoe @hmkyale and I use NHANES data to suggest that the traditional [---] ng/dL threshold may need to be revisited 🧵 1/7" [X Link](https://x.com/aditharun_/status/1919335709862690898) 2025-05-05T10:17Z [---] followers, [--] engagements "Designing antibody drugs depends on developability and binding both now capable of being modeled. Shehata et al [----] [--] report the binding-developability tradeoff which can be incorporated into a global loss function where developability and binding are modeled independently Great new resource for anyone working on antibody CMC developability or discovery. New preprint out today from Ginkgo that benchmarks [---] therapeutic IgGs across [--] developability assays. The resulting dataset is ML-ready and publicly available. If you're thinking about early https://t.co/C9xuHYZOu8 Great new resource for" [X Link](https://x.com/aditharun_/status/1919777457781039105) 2025-05-06T15:33Z [---] followers, [---] engagements "Approved mAbs have melting temps on par with Abs with the highest level of mutations (i.e highest binding affinity). Both have melting temps [--] degrees lower than native Abs so are less stable than native Abs. This suggests that we should be more willing to accept lower developability if it improves binding given we know that approved mAbs operate at this level. Our global loss function should reflect this" [X Link](https://x.com/aditharun_/status/1919777461027360777) 2025-05-06T15:33Z [---] followers, [--] engagements "Improving the individual models for predicting binding and developability along with playing with ensembling method are key. I love the recent open sourcing by @Ginkgo here and companies like @tahoe_ai - itll help progress happen quicker" [X Link](https://x.com/aditharun_/status/1919777463762030698) 2025-05-06T15:33Z [---] followers, [---] engagements "One caveat: Binding affinity is indirectly measured by # of mutations in [--] an imperfect measure (Km would be better) so Im uncertain of the effect size of the developability-affinity tradeoff" [X Link](https://x.com/aditharun_/status/1919777467725701554) 2025-05-06T15:33Z [---] followers, [--] engagements "What model underlies FDAs process here for genAI in these tasks presumably openAIs models given their recent meetings How are they verifying accuracy with their aggressive rollout plan and monitoring for model drift and hallucination Theres not much information on this afaik from jeremy walshs office or makary" [X Link](https://x.com/aditharun_/status/1921239596559724940) 2025-05-10T16:23Z [---] followers, [--] engagements "I think gnomAD puts pathogenic ttr variants at 57ish per 100k so in the world probably 5M ppl w the variants of interest. Assume majority are asx or minimally sx. Variants are enriched in those of African ancestry northern portugal taiwan japan so start w big hospitals in these regions. Assume 1% conservatively have the variant and are asymptomatic in an enriched region. So need to sequence 60k ppl. Cost of targeted sequence assay to detect is [--] usd () so cost is $3M if all goes well" [X Link](https://x.com/aditharun_/status/1922271233426518510) 2025-05-13T12:42Z [---] followers, [--] engagements "Colibactin signature was shown to be positive in 50% of CRC patients younger than [--]. Colibactin signature positivity steadily decreased as age of onset of CRC increased. Only 15% of 70+ year olds were colibactin positive [--]. Why And what does this mean Some thoughts: (1/)" [X Link](https://x.com/aditharun_/status/1923468899179286943) 2025-05-16T20:01Z [---] followers, [--] engagements "Should we care Yes. Colibactin signatures are enriched in early clonal compared to late clonal mutations similar to other well known clock-like signatures that may promote carcinogenesis. Evidence suggests colibactin induces mutagenesis early in life (within the first decade) possibly by APC inactivation [--]. In [--] they presented data that colibactin signatures accounted for a sizeable fraction of APC indel drivers and preferentially did so compared to other genes like p53 which occur later. (2/)" [X Link](https://x.com/aditharun_/status/1923468902257811910) 2025-05-16T20:01Z [---] followers, [--] engagements "If we believe colibactin can inactivate APC we can take pts w positive colibactin signature in their cancer tissue and look at their normal colon tissue. We should expect to see normal crc tissue also have colibactin positive signatures associated with APC changes. Do their normal tissues show this signature If true this would offer more strength to the colibactin idea. (3/)" [X Link](https://x.com/aditharun_/status/1923468906192425319) 2025-05-16T20:01Z [---] followers, [--] engagements "Could it be early exposure to antibiotics causing proliferation of colibactin bacteria C-section Probiotic use with bacteria capable of making colibactin Microplastics Low fiber diets (5/)" [X Link](https://x.com/aditharun_/status/1923468920679235589) 2025-05-16T20:01Z [---] followers, [--] engagements "Low fiber damages colon lining. The association with colibactin positive crc with western diet (low fiber) has been shown [--]. Colibactin producing bacteria thrive in low fiber diets and opportunistically take residence upon damage. With a shift in standard eating habits people may unintentionally consume a low fiber diet increasing odds for CRC. How does low fiber mediate tumorigenesis Is it due to a general inflammatory state or does it allow for some other process like colibactin producing bacteria to act (6/)" [X Link](https://x.com/aditharun_/status/1923468923695186238) 2025-05-16T20:01Z [---] followers, [--] engagements "Other signatures reported in [--] like sbs89 dbs8 and sbs94 were observed preferentially in argentinian and colombian patients. Where are these signatures coming from Argentineans tend to eat more red meat which we know increases odds for crc so maybe these signatures reflect a diet exposure (e.g. red meat low fiber). (7/)" [X Link](https://x.com/aditharun_/status/1923468927356870852) 2025-05-16T20:01Z [---] followers, [--] engagements "Microplastics could damage the colon too leading to pro-tumor state. This has long been a hypothesis in the field. The hard part is nearly everyone in the last [--] decades has been exposed to microplastics making it hard to parse effect. We would also expect a large delay in time between exposure and effect like smoking and lung cancer. But if causal there should be a time and dose relationship (highest microplastic level and longest time under microplastic influence from birth). imagine area under time-dose curve and look at highest quartile of these individuals might offer the largest signal." [X Link](https://x.com/aditharun_/status/1923468934092660865) 2025-05-16T20:01Z [---] followers, [--] engagements "The microplastics CRC question reminds me of the lung cancer smoking connection where the majority of people smoked making it hard to figure things out. Then doll and hill came with their cohort study asking physicians if they smoked then watched the death reports to show the strong relationship. We can do this for microplastics too but we will first need to define microplastic exposure. (10/)" [X Link](https://x.com/aditharun_/status/1923468937280487435) 2025-05-16T20:01Z [---] followers, [--] engagements "I am very interested in trying to understand the reason behind the uptick in CRC in young people. This phenomenon has been highlighted by many including the annual @AmericanCancer report @VinGuptaMD @theNCI and many others. (11/)" [X Link](https://x.com/aditharun_/status/1923468940556087753) 2025-05-16T20:01Z [---] followers, [--] engagements "I agree that risk group is undefined and in an undifferentiated group of patients nnt would be hard to estimate. Being as generous as we can and saying nnt of [--] (based on trials) with a generic statin 300k seems really high. I would have to assume the pcsk9i increase the cost Either way lots of questions with this data. Maybe they count cholesterol screening cost is so low because the lab assay is so cheap to run and then only the exceedingly high values that could lead to pt dying in a year are counted" [X Link](https://x.com/aditharun_/status/1923821376399700432) 2025-05-17T19:22Z [---] followers, [--] engagements "@AdamRodmanMD @MedEdFlamingo What company does the llm detection on multiple choice tests" [X Link](https://x.com/aditharun_/status/1928951226399346842) 2025-05-31T23:06Z [---] followers, [--] engagements "@kenbwork Patients on this tx are going to need to be on immunosupression probably lifelong. Comes with a number of risks. CGM + auto-infusers of insulin to keep A1C and blood glucose titrated seem like better options for patients" [X Link](https://x.com/aditharun_/status/1936891803254632951) 2025-06-22T20:59Z [---] followers, [---] engagements "@Biotech2k1 New to public equity biotech investing do you mean biotech stocks will crash along w rest of market (qqq ath) or separately from rest of market" [X Link](https://x.com/aditharun_/status/1937927273426128919) 2025-06-25T17:34Z [---] followers, [---] engagements "@Biotech2k1 @TohJah7809 Why Just because it has good binding affinity for [--] things so their tox are additive" [X Link](https://x.com/aditharun_/status/1940792241645408701) 2025-07-03T15:18Z [---] followers, [--] engagements "I want to RL some bio stuff and have a few specific problems in mind. What do I read to get up to speed on the RL stuff How do I know if my data has enough inherent information in it to be RL-able Biology is very messy. @willccbb @khoomeik and others DMs welcome" [X Link](https://x.com/aditharun_/status/1942661232185729321) 2025-07-08T19:05Z [---] followers, [----] engagements "Yeah good q. Verifier is straightforward - I have a binary yes/no flag for each sample (so I just need to map the LLM output to 0/1 space maybe cross-entropy). Task is to predict the probability of the outcome happening. The input data for each sample contains 1.n many rows with each row representing a certain point in time before the event happens (some samples may have [--] rows some may have [--] and some may have n in unevenly spaced timepoints). The times across sample are irrelevant. For each time-point for each sample I have p many structured data fields. From this nxp matrix for each sample" [X Link](https://x.com/aditharun_/status/1942693855398355005) 2025-07-08T21:14Z [---] followers, [---] engagements "Verifier is straightforward - I have a binary yes/no flag for each sample (so I just need to map the LLM output to 0/1 space maybe cross-entropy). Task is to predict the probability of the outcome happening. The input data for each sample contains 1.n many rows with each row representing a certain point in time before the event happens (some samples may have [--] rows some may have [--] and some may have n in unevenly spaced timepoints). The times across sample are irrelevant. For each time-point for each sample I have p many structured data fields. From this nxp matrix for each sample I wish to" [X Link](https://x.com/aditharun_/status/1942694087200784704) 2025-07-08T21:15Z [---] followers, [---] engagements "@houndcl I agree that w ur assessment on crdf readout outcome. Naked short is a lot of risk. Any other ways to be synthetically short while keeping risk down given stocks like crdf dont have options markets" [X Link](https://x.com/aditharun_/status/1943029686025515050) 2025-07-09T19:29Z [---] followers, [---] engagements "@MichaelOkun What would the net harm be of giving patients with PD or at high risk for developing it nicotine patches Addiction risk is low and may confer benefit. RCT before doing or collecting more RWE is needed but curious to hear your thoughts" [X Link](https://x.com/aditharun_/status/1944736507727601959) 2025-07-14T12:31Z [---] followers, [--] engagements "The majority (60%) of genetic diseases in hospitalized infants are single nucleotide substitutions. Base editing ($BEAM) as accomplished by Kiran Musunuru and team for Baby KJ and prime editing ($PRME) can fix these. We have to 1) sequence infants with suspected genetic disorders and figure out IF the variant is actually harmful 2) platform N-of-1 gene editing treatments Of the [---] variants identified in infants with a suspected genetic disorder by either WGS or targeted sequencing [---] were single nucleotide substitutions in a [----] JAMA study. The hard part is figuring out if a variant is" [X Link](https://x.com/aditharun_/status/1945232962036826143) 2025-07-15T21:24Z [---] followers, [---] engagements "@SinaHartung Have you tried using the open evidence API to build Eureka" [X Link](https://x.com/aditharun_/status/1946224864102174824) 2025-07-18T15:05Z [---] followers, [--] engagements "@nikillinit The company that serves the LLM needs a ad marketplace where others can bid and buy mentions similar in spirit to current ad search methods. That said google is probably in the best place to do this well" [X Link](https://x.com/aditharun_/status/1947326403478786312) 2025-07-21T16:02Z [---] followers, [--] engagements "How can we accurately classify cause of death in death certificates Data on US mortality rates and trends for specific causes of death including my own work comes from the cause(s) of death listed on the death certificate. And these reports are largely filled out by overworked residents and doctors who are trying to figure out what happened to a patient who died out of the hospital in between notes/rounding/clinic/other duties. This is what I see in the hospital and @adamcifu had the same experience [--]. Take for example this paper [--] showing 51% of death certificates had major errors. Expert" [X Link](https://x.com/aditharun_/status/1947358808327479747) 2025-07-21T18:11Z [---] followers, [---] engagements "* The current medical system does not compensate physician time to adjudicate cause of death. It is important work though. This is a great opportunity for an LLM to step in. AI can read the chart offer its guess as to cause of death as a second data point on the patient's cause of death. Researchers can take this extra data point into account in their work" [X Link](https://x.com/aditharun_/status/1947358809971683339) 2025-07-21T18:11Z [---] followers, [--] engagements "@dylan522p Is this Taiwan equivalent ISEF or a regular high school exhibition Would be wild if latter" [X Link](https://x.com/aditharun_/status/1947746017488867723) 2025-07-22T19:50Z [---] followers, [----] engagements "$srpt's viral vector appears to trigger liver failure in a small subset of patients that led to [--] deaths. And the 3rd death in the limb girdle is probably related. What immunosuppression regimen can $srpt use to prevent immune cascade Do we have to immunosupress everyone What supporting therapies can we use to prevent acute liver failure if even mild liver enzyme spikes are observed How can we identify which minority of patients will experience serious liver injury The viral vector is the common thread here. It is reasonable that FDA revoked platform designation for $srpt viral vector in" [X Link](https://x.com/aditharun_/status/1947820457685663808) 2025-07-23T00:46Z [---] followers, [---] engagements "* From what I gather: 2d from infusion: viral capsid is presented on hepatocyte MHC1 and viral dna sensed by innate immune system (liver is main sink for $srpt viral vector). Innate immune system triggers cytokine release and local immune reaction causing small vessel thrombosis / angiopathy. few weeks: T cells against the capsid proliferate and return to liver 1-2 months: T cells attack capsid presenting hepatocytes (liver enzymes flare) 2-3 months: depending on extent of damage from thrombosis and hepatocyte death patients can progress to liver injury" [X Link](https://x.com/aditharun_/status/1947820459636019392) 2025-07-23T00:46Z [---] followers, [---] engagements "@john_hersc79276 Agreed that fighting FDA is not a good strategy. The goal is to work together" [X Link](https://x.com/aditharun_/status/1947820873185886434) 2025-07-23T00:47Z [---] followers, [--] engagements "@princetongb Wow. compared to viking having [--] fte" [X Link](https://x.com/aditharun_/status/1948414432284913997) 2025-07-24T16:06Z [---] followers, [--] engagements "$CRDF observed in a post-hoc Ph2 analysis that bevacizumab-naive KRAS mutant mCRC patients (n=13) treated with bevacizumab+FOLFIRI+onvansertib had a median PFS of [----] and ORR of 77%. They state that the "exceptional efficacy observed in our study in bevacizumab-nave patients is unique to onvansertib". And later the post-hoc analysis "underscores the unique sensitivity of this population to onvansertib. This exceptional response alongside our translational studies showing onvansertibs effect on the hypoxia pathway and angiogenesis supports shifting onvansertib development to the frontline" [X Link](https://x.com/aditharun_/status/1948489498792182203) 2025-07-24T21:04Z [---] followers, [----] engagements "Nothing rubs me the wrong way more than a friend venmo requesting $4.18 for my share of the six pack he bought" [X Link](https://x.com/aditharun_/status/1949116745374994741) 2025-07-26T14:36Z [---] followers, [--] engagements "The technology underlying companion diagnostics determine who can receive the drug. It dictates who can even have a chance at realizing benefit from the drug. The major inclusion criteria for adagrasib is a KRAS G12C mutation but executing to detect this mutation can leave patients out. Here are some ways: WGS at 80x coverage of tumor identified KRAS G12C mutations in [--] of [--] lung adenocarcinoma patients previously labelled as RAS/RTK/RAF negative from WES [--]. These patients would have previously not been eligible for adagrasib/sotorasib but could have benefited if sequencing had identified" [X Link](https://x.com/aditharun_/status/1950232997577335264) 2025-07-29T16:32Z [---] followers, [---] engagements "The companion diagnostic for adagrasib [--] uses a patient's cfDNA sequenced on Illumina NovaSeq6000 to detect KRAS G12C mutations if the log likelihood score is above a threshold. The log likelihood score is based on mutant allele fraction (variant reads / total reads) and unique variant molecules to reflect the likelihood of the variant being true signal over noise. If we use MAF from DNA isolated from WBCs or a matched normal we could determine if the cfDNA variant is background or not with more sensitivity" [X Link](https://x.com/aditharun_/status/1950232999666372977) 2025-07-29T16:32Z [---] followers, [--] engagements "@Biotenic I only saw [--] reddit post couldnt find any others which made it tough to assess" [X Link](https://x.com/aditharun_/status/1950990943697457428) 2025-07-31T18:44Z [---] followers, [--] engagements "A [--] nucleotide deletion (delGT) in NCF1 which encodes p47 a subunit of NADPH oxidase renders NADPH oxidase defective. This prevents neutrophils from generating reactive oxygen species. Patients get recurrent bacterial and fungal infections typically catalase-positive bugs. Stem cell transplant from a donor is the only potential cure. Enter prime editing. $PRME uses prime editing to correct the delGT and restore neutrophil NADPH oxidase activity. It's an ex vivo solution similar to Casgevy - a patient's stem cells are collected the prime editing system is added and the edited cells are" [X Link](https://x.com/aditharun_/status/1950992681800187938) 2025-07-31T18:51Z [---] followers, [---] engagements "Beam licenses Acuitas LNPs the same class of LNPs Verve uses. Prime uses their own GalNAc-LNP. LNPs can have significant safety events so will be interesting to see how Prime's own LNP performs. Also what type of LNP will Prime use for their CF program SORT LNPs are a good choice imo. Lipid chemistry is fascinating" [X Link](https://x.com/aditharun_/status/1951029338893328711) 2025-07-31T21:16Z [---] followers, [---] engagements "@alamentarius How does mfn change some of this Not sure how to think about pricing" [X Link](https://x.com/aditharun_/status/1951049355114291516) 2025-07-31T22:36Z [---] followers, [---] engagements "@BalaBioResearch Perhaps market feels execution of sales insurance coverage or executive team operating capability is doubted. Not sure something to investigate any thoughts" [X Link](https://x.com/aditharun_/status/1951793891138548173) 2025-08-02T23:55Z [---] followers, [---] engagements "The biology is interesting. TL;DR instead of fixing the issue in AGXT they try to prevent the precursor of the toxic compound from being created by introducing more mutations into a different gene (HOA1). The substrate reduction approach. Primary hyperoxaluria Type [--] (PH1) comes from mutations in AGXT that turns off alanine-glyoxylate aminotransferase activity. This forces glyoxylate to buildup and be converted into oxalate. Oxalate accumulation causes symptoms (kidney stones growth delay bone pain anemia cardiac issues non-healing ulcers). Arbor approaches PH1 not by trying to fix AGXT but" [X Link](https://x.com/aditharun_/status/1952034245259464992) 2025-08-03T15:50Z [---] followers, [---] engagements "Best-in-disease may be a little premature but $PRAX vormatrigine results are good very comparable to xenon. Patients w [--] seizures on average [--] anti-seizure meds at baseline and vormatrigine has safety profile on par with other anti-epileptics (worse than xenon) + 56% median reduction in seizure burden at week [--] + 22% are at zero burden during observational period. Safety is a little bit worrying compared to xenon data but otherwise great results" [X Link](https://x.com/aditharun_/status/1952365641412837837) 2025-08-04T13:46Z [---] followers, [----] engagements "Notes from $PRAX call this AM: Vormatrigine d/c rate due to improper titration of background ASMs. Going to be more deliberate in dose reduction of background ASMs No food effect w vormat. Xenon has with food label. Xenon has once daily oral dosing w no titration like vormat. But half life much longer (24 vs 6ish). Most patients are women. Xenon and prax drugs so far not noted to cause problems with contraception but long term data needed. Prax has no cyp induction but xenon asks that barrier method be used if patient on hormonal only. Potential advantage praxis here. Adding [--] mg vormat arm" [X Link](https://x.com/aditharun_/status/1952484853326471268) 2025-08-04T21:40Z [---] followers, [---] engagements "$XENE excluded cenobamate in Phase 2b and [--]. $PRAX does not shy away in cenobamate group (iirc 30% of patients on hefty doses) 55% median reduction in seizure burden. I would be very interested in seeing the AE stratified into cenobamate users versus not" [X Link](https://x.com/aditharun_/status/1952488654842347841) 2025-08-04T21:55Z [---] followers, [---] engagements "Power-1 readout w placebo will be key for evaluating vormatrigine. August conference is going to be more $prax radiant data I think. That said I see no reason the science will not work. AEs and deepening of responses are two of the most salient concerns. AEs would be great if prax did subset by cenobamate because they kept saying that on the call w no data shown to back it up. If you believe this + background asm titration being mismanaged then its fine. Im not a neuro/psych doc so would love some input here. Xenon didnt report the same concern w background asm but they also didnt have" [X Link](https://x.com/aditharun_/status/1952908906272874567) 2025-08-06T01:45Z [---] followers, [---] engagements "PURE speculation: Head of hypoimmune program at $SANA (Sonja) left to work on research and questions she is really curious about (and has always been). Perhaps she is not particularly interested in seeing the rest of the clinical studies / execution (compared to opportunity cost of working on hypoimmune organs and other research questions) if a lot of it is "solved" and many other MDs have the knowledge/experience to see it through well. She started as a Professor at Cedars-Sinai in Aug [----] along w appointment at SANA per LinkedIn. So perhaps she agreed/wanted to stay on till the initial" [X Link](https://x.com/aditharun_/status/1953245957728157739) 2025-08-07T00:05Z [---] followers, [----] engagements "Borzoi is also directionally correct at identifying expression changes of a SNP. AUROC 0.8+ and AUPRC 0.9+ when SNP is within 3kb of TSS. However it does not do a good job at estimating effect size of SNP (see panel d) w/ Spearman correlation of [---] and doesn't tell you how rare a variant is (as we expect from modeling setup)" [X Link](https://x.com/aditharun_/status/1953506675413172548) 2025-08-07T17:21Z [---] followers, [---] engagements "RFDiffusion can make a high-affinity (pretty) specific de novo binder from just the target protein sequence. Feed in sequence of protein you want to bind to -- RFDiffusion makes designs -- Protein MPNN fleshes out sequence of binder -- AF3 folds and filters based on how plausible the pose is (you can combine with binding affinity predictors too) -- experimentally synthesize and test binding affinity. This goes beyond the functional motif scaffolding problem where you know a binding motif and want to optimize that in a future design. RFDiffusion does that too but the hard part is knowing a" [X Link](https://x.com/aditharun_/status/1953799594128585151) 2025-08-08T12:44Z [---] followers, [----] engagements "More than twice the number of Oreos sold in [----] are sold this year despite 12% of Americans having tried a GLP-1 (6% currently on it). [--] billion Oreos were sold in the US in [----]. Today over [--] billion Oreos are sold. Fascinating. Oreo sales keep going up even though a lot of people have taken weight loss drugs. The Oreo substitution diet begins" [X Link](https://x.com/aditharun_/status/1955050289649582421) 2025-08-11T23:34Z [---] followers, [---] engagements "Great writeup Two thoughts 1) what if we dosed organic lithium well below the human-equivalent dose from mice studies to try to ward off Li-tox AEs would it do anything 2) serum Li is not equal to brain Li. What chemistry tricks can we do (packaging/targeting/composition) to shift more to brain" [X Link](https://x.com/aditharun_/status/1955744104001761419) 2025-08-13T21:31Z [---] followers, [--] engagements "After selling out not too long ago burry has gone bullish on united healthcare asml regeneron handful of chinese companies and a few other names. I did not expect burry to go long when market is roaring A lot of people are calling market top atm with ath in spy/qqq AI revenues being a product of vc paying indirectly for compute inflation etc BREAKING 🚨 Michael Burry just bough United Health too He just disclosed owning up to $100M of calls on $UNH https://t.co/vrfVbKpLq4 BREAKING 🚨 Michael Burry just bough United Health too He just disclosed owning up to $100M of calls on $UNH" [X Link](https://x.com/aditharun_/status/1956176403503681985) 2025-08-15T02:09Z [---] followers, [---] engagements "What happened with $PGEN has happened many times before for other orphan drugs. This is not a reflection of the current administration or Vinay Prasad. Seeing a lot of things about Prasad's quote shown below. He is right for rare severe diseases where no decent treatment currently exists. And this is not something that is novel its been done by previous directors / administrations too. Prasad has tweeted things like "RCT or STFU" but this stance is one that directors before him have taken and reflect the consensus opinion of drug evaluators in rare diseases" [X Link](https://x.com/aditharun_/status/1956389583626621030) 2025-08-15T16:16Z [---] followers, [---] engagements "Is $REPL going to get approved given $PGEN decision by Prasad" [X Link](https://x.com/aditharun_/status/1956755637305815106) 2025-08-16T16:31Z [---] followers, [---] engagements "In @A_May_MD pitch on $abvx he points out obefazimod inverse dose relationship. It should not have made people skeptical for [--] reasons: 1) rna dynamics make this plausible 2) in protein-receptor space this is true of inverse agonist drugs. Ofc hindsight is 20/20" [X Link](https://x.com/aditharun_/status/1956885683257405706) 2025-08-17T01:07Z [---] followers, [----] engagements "The survival curves of $ELTX mutant KRAS vaccine as adjuvant therapy in PDAC published in @NatureComms skim the median. The OS stays just above [--] from [--] to [--] months and a single event occurs dropped OS to zero (after [--] patients are censored). A similar story occurs with panel B where relapse free survival hovers around 60% from [--] to [--] months (2 patients get censored) one event moves it below 50% and the remaining patients get censored" [X Link](https://x.com/aditharun_/status/1957496192687824979) 2025-08-18T17:33Z [---] followers, [---] engagements "@varma_ashwin97 @DrSamuelBHume Clever hadnt heard of this before. Makes sense though. Could get pio approved if obesity-related endpoints are flat (mace obesity-related cancer). Obviously pio helps a1c" [X Link](https://x.com/aditharun_/status/1957875802227257691) 2025-08-19T18:42Z [---] followers, [--] engagements "People know glp1 = nausea etc. so if they dont have then they want to drop out to get drug/other glp1 perhaps some other way. How to drop out Maybe state AE and leave Were these marker/imaging/clinician witnessed (like you say you are having diarrhea but did I see it physically happen this is outlandish but captures the vibe) I have no position in viking. If we assume the above to be true then it raises the following point - its tough to evaluate any placebo arm ever and we can always read the placebo tea leaves however we want" [X Link](https://x.com/aditharun_/status/1957907637737082904) 2025-08-19T20:48Z [---] followers, [--] engagements "@AnilMakam Metformin and GLP1 curves are basically the same right So not magic just that metformin and glp1 both superior to insulin. Not mechanistic difference because insuling tx patients have more severe disease and other RF for cancer" [X Link](https://x.com/aditharun_/status/1958924680699813960) 2025-08-22T16:10Z [---] followers, [---] engagements "$IMRX up 65% since this tweet 4d ago Can dual MEKi and mutant KRASi add value over KRASi in PDAC (virtually all PDAC has a KRAS mutation) Atebimetinib (orphan drug status) is MEKi ($IMRX) and many KRASi approaches are approved (Sotorasib and Adagrasib) or in the works (Revolution Medicines). Would it make sense https://t.co/htKA8zUCkG Can dual MEKi and mutant KRASi add value over KRASi in PDAC (virtually all PDAC has a KRAS mutation) Atebimetinib (orphan drug status) is MEKi ($IMRX) and many KRASi approaches are approved (Sotorasib and Adagrasib) or in the works (Revolution Medicines). Would" [X Link](https://x.com/aditharun_/status/1960010221637804215) 2025-08-25T16:03Z [---] followers, [---] engagements "$PRAX is mispriced it should be worth at least 3B in market cap. Currently it sits at 1B market cap. Why Ulixacaltamide drug for essential tremor won't work but that's fine. Assign zero value to that asset. Relutrigine will work for SCN2A and SCN8A epilepsy but ALSO broad DEE too. Assume [------] total patients 20% max penetration slow linear ramp to peak penetration over [--] years $10k/pt/year discount rate at 5% annually. NPV of [--] years of sales is 3.6B. But that's not all. They have elsunersen which treats an ultra-rare disease of SCN2A GoF epilepsy patients. Conservatively throw a 200M NPV" [X Link](https://x.com/aditharun_/status/1960815018028753118) 2025-08-27T21:21Z [---] followers, [----] engagements "@KobeissiLetter Stock buyback at 4T valuation 🤯 Jensen feels they are undervalued at this point in the AI cycle" [X Link](https://x.com/aditharun_/status/1960817254511067608) 2025-08-27T21:30Z [---] followers, [----] engagements "@anssiapina 1) ok interesting then pts can recieve tx immediatet 2) agreed But does the extension of life come with any symptom improvement. I think either way its a choice for each pt to make based on their desires/wants/needs" [X Link](https://x.com/aditharun_/status/1961064130615337021) 2025-08-28T13:51Z [---] followers, [---] engagements "Two books I've been meaning to read for a while: 1) george soros: a life in full and 2) the alchemy of finance" [X Link](https://x.com/aditharun_/status/1961087333278642541) 2025-08-28T15:23Z [---] followers, [---] engagements "@ktkadakia @bostonsci @JACCJournals Woah incredible stuff Great to see boston scientific take into consideration research to make changes to their products" [X Link](https://x.com/aditharun_/status/1961535358304202755) 2025-08-29T21:04Z [---] followers, [--] engagements "@SinghJyotirmai What about the tariffs though How did US tariff announcement change India-China trade" [X Link](https://x.com/aditharun_/status/1961540346522808813) 2025-08-29T21:23Z [---] followers, [--] engagements "Sangamo ($SGMO) needs funding to continue. They have like [--] months left. They are coming to Wells Fargo and Cantor conferences in the next few days because they are either finalizing a deal or they want to drum up deal interest. I'm not sure which way its going to go. The thing tipping me to lean 55-45 finalizing a deal is that they have good Fabry data and they are still hiring actively. Dead companies don't hire. I could be wrong and this will serve as a public scorecard" [X Link](https://x.com/aditharun_/status/1963062554692845682) 2025-09-03T02:12Z [---] followers, [---] engagements "@princetongb What do you recommend I read to orient myself to macro stuff. I don't understand it well" [X Link](https://x.com/aditharun_/status/1963308260389167159) 2025-09-03T18:29Z [---] followers, [--] engagements "@RNAiAnalyst Im just confused why they keep opening job positions. Is it a headfake" [X Link](https://x.com/aditharun_/status/1963683758676292029) 2025-09-04T19:21Z [---] followers, [--] engagements "(5/) Areas for improvement Data data data High-quality multi-modal data curated for the problem at hand is the key. Architecture-wise how do we capture long distance informatio losing local detail Is context length the answer. Probably not. Local signal degrades as context length scales. And 1M nucleotides (largest length to date) is not even near chromosome level yet. Optimization environment: Can we set up a reinforcement learning environment for biology a la Prime Intellect Hub. Benchmarks: we need consensus datasets + metrics. Each DNA FM paper evaluates on different datasets and uses" [X Link](https://x.com/aditharun_/status/1963730922945007935) 2025-09-04T22:28Z [---] followers, [--] engagements "(6/) Applications of DNA Models: [--]. Designing state-specific promoters for CAR T AAV gene therapy and ddRNAi drugs [--]. Discovering new disease-causing targets through in silico mutagenesis [--]. Resolving variants of unknown significance [--]. Biosecurity These ideas are explored here: https://www.aditharun.com/p/dna-foundation-models https://www.aditharun.com/p/dna-foundation-models" [X Link](https://x.com/aditharun_/status/1963730925554180446) 2025-09-04T22:28Z [---] followers, [---] engagements "I think the ACMG guidelines are reasonable (Richards et al [----] Genetics in Medicine This evidence framework is what Ambry/Foundation/Myriad and others are judged against to afaik. They have a section on in silico prediction tools and write "These are only predictions however and their use in sequence variant interpretation should be implemented carefully. It is not recommended that these predictions be used as the sole source of evidence to make a clinical assertion". I agree DNA models or any other such model is a prediction at the end of the day. Table [--] attached is their classification of" [X Link](https://x.com/aditharun_/status/1964403017416941682) 2025-09-06T18:59Z [---] followers, [---] engagements "Im noticing weird behavior in options bids in low volume small market cap stocks. I think its algo trading / bot behavior. I place bid slightly above current bid I refresh and I am only person with said bid since it shows volume of [--] at that price [--] mins later I see [--] other bids at that higher price I cancel my bid now within 30s i see [--] bids spread across multiple premiums all slightly below my cancelled bid" [X Link](https://x.com/aditharun_/status/1966539651843264900) 2025-09-12T16:29Z [---] followers, [---] engagements "@GeneInvesting @Clarkitect80 From exhibit [----] filed w SEC in 2020" [X Link](https://x.com/aditharun_/status/1966547357274624233) 2025-09-12T17:00Z [---] followers, [--] engagements "@GeneInvesting @Clarkitect80 Plus minus a B or two" [X Link](https://x.com/aditharun_/status/1966551211760300433) 2025-09-12T17:15Z [---] followers, [--] engagements "It is incredible that GLP1s work for reducing cravings with evidence in alcohol use disorder opioid use disorder and cannabis use disorder. A little peptide that comes from the proglucagon gene primarily made by pancreatic and gut cells (and to a lesser extent in the brain) works like magic. It also fixes type [--] diabetes prevents heart attacks corrects excessive weight decreases incidence of many cancers and Alzheimer's disease. A GLP1-like medicine for preventing the habit circuit that leads to endless social media consumption would be incredible. Downstream consequences of accomplishing" [X Link](https://x.com/aditharun_/status/1966614705876693303) 2025-09-12T21:27Z [---] followers, [----] engagements "@3primeAnalytica @chris342222 What do you think about their value heading into rate cuts and possible inflation" [X Link](https://x.com/aditharun_/status/1967653305762660783) 2025-09-15T18:14Z [---] followers, [--] engagements "$qure however relies on AAV to treat Huntingtons Exclusive: Biogen is exiting AAV gene therapy joining a growing list of pharmas stepping back from the costly risk-laden delivery tech. https://t.co/lTVmPzciLS Exclusive: Biogen is exiting AAV gene therapy joining a growing list of pharmas stepping back from the costly risk-laden delivery tech. https://t.co/lTVmPzciLS" [X Link](https://x.com/aditharun_/status/1971742845108474356) 2025-09-27T01:05Z [---] followers, 12K engagements "ALS drug seemed to have a substantial OS benefit when comparing to external propensity-matched control arm (baseline characteristics pretty similar) but any benefit disappeared when comparing to placebo in intention-to-treat. This is important given the recent news about $QURE. They showed massive treatment benefit in Huntington's disease patients compared to an external control arm. This effect may disappear with placebo (or it may not and that's wonderful for patients) Note that $QURE's high-dose treatment arm that showed efficacy was n = [--] versus the ALS trial treatment arm that had n =" [X Link](https://x.com/aditharun_/status/1972016473850937732) 2025-09-27T19:12Z [---] followers, [----] engagements "OV329 does not accumulate in brain So if you give [--] mg/kg/d of the $ovid drug it never accumulates in the brain.so any on-target activity is transient inhibition Per poster abstract "Brain retina eyeball and blood samples were collected at [--] hours after initiation of SC infusion." Maybe I've missed something obvious here" [X Link](https://x.com/aditharun_/status/1972794990956105730) 2025-09-29T22:45Z [---] followers, 14.8K engagements "@JusticeforB_MD @ElicioTx Agreed Competitive landscape of 1L and 2L pancreatic cancer is poised to change soon $imrx meki in metastatic pancreatic cancer $rvmd pan-kras codon [--] inhibitor in metastatic pdac $eltx kras amphiphile vaccine in locoregional" [X Link](https://x.com/aditharun_/status/1972858926011150426) 2025-09-30T02:59Z [---] followers, [---] engagements "Has good news been announced for any public biotech company when they guided for a specific quarter missed that deadline completely and did not put out any release/commentary offering an explanation or postponement Re: $ovid q3 ph [--] topline" [X Link](https://x.com/aditharun_/status/1973460390727262382) 2025-10-01T18:49Z [---] followers, [----] engagements "Announced today: Evolocumab (PCSK9i) met both primary endpoints (time to first CV event time to CV event or ischemia-driven arterial revascularization) in VESALIUS-CV. Amgen positioning PCKS9i + lipid lowering therapy as the basis for primary prevention of CV disease. This trial can help inform future strategies for primary prevention of CV disease. Evo is given subq q2/q4 weeks so opportunity exists for a therapy that needs to be given less often. ANGPTL3 inhibition and remnant cholesterol targeting drugs could play a role. Atherosclerotic CV disease often co-occurs with other chronic" [X Link](https://x.com/aditharun_/status/1973942114850648326) 2025-10-03T02:44Z [---] followers, [----] engagements "@TCellEngager @jesse_brodkin If EMA rejects then I'm not sure what the company has. Schizophrenia pipeline doesn't make sense Fragile X MoA is totally distinct from Sigma-1R so that won't work etc. But then again stock went up 10% today on schizophrenia data announcement. so im confused" [X Link](https://x.com/aditharun_/status/1974269888270323753) 2025-10-04T00:26Z [---] followers, [---] engagements "I need to think more about the pdac clinical landscape. I like their approach and data has been clean in that no tricks ive discovered on reporting information. Small n which is the main concern but that will be solved w ph3. The extent to which low dose chemo might be driving the signal is another concern before I reach a conclusion. Fwiw rn i do not think low dose can have this sort of effect" [X Link](https://x.com/aditharun_/status/1975002169633083806) 2025-10-06T00:56Z [---] followers, [--] engagements "@hannibalspeaks sports betting but better" [X Link](https://x.com/aditharun_/status/1977493054782144633) 2025-10-12T21:54Z [---] followers, [--] engagements "What's a good book to read on Lee Kuan Yew" [X Link](https://x.com/aditharun_/status/1977812650567623135) 2025-10-13T19:04Z [---] followers, [---] engagements "Nicotine promotes a chronic inflammatory state promote pro-inflammatory cytokine release and suppresses immune responses against Gram-negative bacteria by impairing mitrochondrial function. This acts as a nidus for periodontal disease gum recession and tooth loss caused primarily by gram-negative anaerobic bacteria. Nicotine-free pouches (like Ultra) try to deliver the same hit without the consequences of nicotine. There is no free lunch but maybe its a good discount lunch. I have no idea I've never tried them. But here's what's in them: [--]. Paraxanthine Paraxanthine (17-dimethylxanthine) is" [X Link](https://x.com/aditharun_/status/1978072882308681945) 2025-10-14T12:18Z [---] followers, [---] engagements "I had a toilet issue today. Normally Id call someone but this time I put in my AirPods opened ChatGPT voice mode and walked through it step by step. I reasoned through what I was seeing what the parts seemed to do and what I thought should happen with the AI. I had zero experiencejust knew there was a chain and a fill valve. But with voice mode and occasionally flipping to video I got 90% of the way there. The last 10% I figured out myself. The big thing is that ChatGPT voice allowed me to do something that I wouldve never felt comfortable doing before" [X Link](https://x.com/aditharun_/status/1978247321360928962) 2025-10-14T23:51Z [---] followers, [---] engagements "Im really intrigued about holding $jepi or similar high yield equities. The primary risk is they cut dividends when cash flow drops" [X Link](https://x.com/aditharun_/status/1978906836674707840) 2025-10-16T19:32Z [---] followers, [---] engagements "The states w highest median income tend to have higher credit card debt Based on eyeballing the data fyi Americans are drowning in credit card debt: US household credit card debt hit a record $1.33 trillion in August. The combined average credit card balance across all age groups is now at whopping $10668 PER household. States with the largest average debts in Q2 [----] were https://t.co/QSZgTBa8XB Americans are drowning in credit card debt: US household credit card debt hit a record $1.33 trillion in August. The combined average credit card balance across all age groups is now at whopping" [X Link](https://x.com/aditharun_/status/1978967121670381747) 2025-10-16T23:31Z [---] followers, [---] engagements "@RyanSTTrader I dont understand this. Spy is flat on the week and sentiment is extreme fear. I guess forward looking anticipation of drop is the consensus" [X Link](https://x.com/aditharun_/status/1979208470957015439) 2025-10-17T15:30Z [---] followers, [--] engagements "Pregnant women with HIV on treatment and with CD4 [---] may have significantly lower odds of having a preterm birth if given tmp-smx during pregnancy. 2% of births in the trimethoprimsulfamethoxazole group were preterm as compared with 14% in the placebo group. The table below shows the results for HIV positive women in this study - better in many categories when given tmp-smx (cotrimoxazole). These are exploratory analyses not powered to make claims. We need an RCT specifically looking at this population. This study finds that pregnant women given TMP-SMX 80mg-400mg BID after [--] weeks" [X Link](https://x.com/aditharun_/status/1932104457149636666) 2025-06-09T15:56Z [----] followers, [---] engagements "Google has an incredible amount of data to leverage for LLM development. They fine-tuned a LLM for CRISPR-GPT based on [--] years of a public Google Discussion Group on CRISPR gene-editing. Now of course they can't use X which is probably the best real-time source of information. Interestingly real-world research questions performance drops 20% when adding in this source of data. Perhaps the fine tuned model over-indexes on the forum which contains a lot of wrong deep technical CRISPR questions (because scientists post what they cannot figure out / confused about)" [X Link](https://x.com/aditharun_/status/1952884072453423346) 2025-08-06T00:07Z [---] followers, [---] engagements "Tbh I thought $vyne [---] results would be good. I was surprised when they were so bad. But I think its something to dissect and learn from. On my todo list. A post-mortem on this could be very informative" [X Link](https://x.com/aditharun_/status/1953063662136041662) 2025-08-06T12:00Z [---] followers, [---] engagements "Zero shot Evo2 (AUROC 0.65ish) vs. ProteomeLM-Ess fine-tuned for essentiality (AUROC 0.9) for predicting gene essentiality. However ProteomeLM-Ess is only for protein coding genes. What is Evo2 zero-shot performance subset to only protein coding genes With fine-tuning Evo2 can be as good as ProteomeLM-Ess" [X Link](https://x.com/aditharun_/status/1953171719021027426) 2025-08-06T19:10Z [----] followers, [---] engagements "Can dual MEKi and mutant KRASi add value over KRASi in PDAC (virtually all PDAC has a KRAS mutation) Atebimetinib (orphan drug status) is MEKi ($IMRX) and many KRASi approaches are approved (Sotorasib and Adagrasib) or in the works (Revolution Medicines). Would it make sense to combine the two. We can look to Sotorasib (KRAS G12Ci) + Trametinib (MEKi) in heavily pre-treated NSCLC patients (median [--] lines of prior tx) had ORR of 14% (prior KRAS G12Ci given n = 7) and 9% (KRAS G12Ci naive n = 11). In CRC patients ORR of [--] in prior KRAS G12Ci (n = 3) and 20% in KRAS G12Ci naive (n = 15)." [X Link](https://x.com/aditharun_/status/1958220089674580275) 2025-08-20T17:30Z [----] followers, [----] engagements "Yeah. All patients had no cancer at start and some developed cancer. If we think about CRC the HR [---] 0.46-0.64 doesn't really make sense. They propensity matched on a number of criteria including adverse socioeconmic status and prior colonoscopy. We know T2DM patients requiring insulin have more severe disease than those who do not require insulin. We have no idea what the A1C levels are between groups in this study. We also know routine colonoscopies can detect and treat polyps that lead to CRC. We do not know if patients with T2DM on insulin had different colonoscopy rates following" [X Link](https://x.com/aditharun_/status/1958935231941001423) 2025-08-22T16:52Z [----] followers, [--] engagements "Sangamo is an interesting company. Summary of their work and my thoughts on their specific clinical programs. 1) They have a brain specific AAV that delivers zinc fingers (ZF) partnered with Lilly/Genentech/Astellas. It has not been tried in humans yet only primates. But in primates it seems to work. If partner co's can't find something good no further milestone payments. I expect this to be the case. 2) They have primate data on IV ZF + brain AAV in prion disease. I'm skeptical of prion results (see reply). And $SGMO has an intrathecal ZF + brain AAV product for idiopathic small fiber" [X Link](https://x.com/aditharun_/status/1960922188799692905) 2025-08-28T04:27Z [---] followers, [----] engagements "$SGMO ASGCT talk presented prion primate data. In primates (n=2) prion transcript 25% of full disease in different brain regions. Does 75% reduction in NEW prion transcript matter Patients with prion disease progress quickly receive testing for all sorts of other things then get ruled into this diagnosis. So by the time patient arrives ready for treatment they will have lots of prion protein already misfolded and aggregated. This drug slows rate of NEW misfolded proteins joining the aggregated by 75%. Will that translate into regression of patient symptoms or even stabilization Probably not" [X Link](https://x.com/aditharun_/status/1960922191341461511) 2025-08-28T04:27Z [----] followers, [----] engagements "@yusufhameed @MofeekS Appreciate the viewpoint from pain perspective. Did not know" [X Link](https://x.com/aditharun_/status/1961080862562218042) 2025-08-28T14:58Z [----] followers, [--] engagements "This is really cool Interesting that all post-tx muts were present pretx suggesting the tx acts exerts selection pressure but is not a wbc mutagen. Do you think it becomes a mutagen if chemo given for many more cycles If Im interpreting the vaf change per month plot on each tx at [--] year a 12% median increase in vaf is expected on chemo What % increase is considrered pos selection in your km curves" [X Link](https://x.com/aditharun_/status/1963228282729496996) 2025-09-03T13:11Z [----] followers, [--] engagements "Focal onset epilepsy has a few key players at the moment ( $RAPP $XENE $PRAX) but it is my belief that $PRAX is undervalued and has a lot of upside. Described my thoughts here: If $PRAX executes on rare disease epilepsy alone it will justify a 3B valuation imo (3x upside from here) and I feel that is almost a sure thing. https://www.aditharun.com/p/praxis-epilepsy-drugs-will-work-but https://www.aditharun.com/p/praxis-epilepsy-drugs-will-work-but https://www.aditharun.com/p/praxis-epilepsy-drugs-will-work-but https://www.aditharun.com/p/praxis-epilepsy-drugs-will-work-but" [X Link](https://x.com/aditharun_/status/1969460517551935774) 2025-09-20T17:55Z [----] followers, [---] engagements "@adamcifu illustrating Goodhart's Law in medical school admissions "Those uninterested in research sacrifice time during which they could be doing what they love to fill third-tier journals with drivel" Goodharts Law and Medical School Admissions Sensible Medicine Oct [--] 2025" [X Link](https://x.com/aditharun_/status/1977000106810654934) 2025-10-11T13:15Z [---] followers, [----] engagements "$PRAX up 224% today The valuation pre-jump today supported a 2-3x upside just off epilepsy portfolio and it came with a flyer on the essential tremor drug And today they announced Positive Topline Results from Two Pivotal Phase [--] Studies of Ulixacaltamide HCl in the Essential3 Program for Essential Tremor I detailed this in the quoted post Have not studied the data or results will look into it. My initial thoughts are that Im quite surprised the ET trial was a success $PRAX is mispriced it should be worth at least 3B in market cap. Currently it sits at 1B market cap. Why Ulixacaltamide drug" [X Link](https://x.com/aditharun_/status/1978827929674002710) 2025-10-16T14:18Z [---] followers, [----] engagements "Regeneron 👀 on the heels of NEJM DB-OTO results The [--] selected: Makary says more vouchers set to be announced in the coming months as well. https://t.co/bOxlnTuiE9 The [--] selected: Makary says more vouchers set to be announced in the coming months as well. https://t.co/bOxlnTuiE9" [X Link](https://x.com/aditharun_/status/1978947039728713801) 2025-10-16T22:11Z [----] followers, [---] engagements "Lp(a) drugs are coming soon Pelacarsen (Ionis) Lepodisiran (Eli Lilly) Olpasiran (Amgen) Zerlasiran (Silence Therapeutics) All with positive Phase [--] and in phase 3; All subq injections Muvalaplin (Eli Lilly) Positive Phase [--] announced. Oral route of administration. It's happening Lipoprotein(a) testing is going to skyrocket once it's possible to treat it. https://t.co/aFhMiA4Zu6 It's happening Lipoprotein(a) testing is going to skyrocket once it's possible to treat it. https://t.co/aFhMiA4Zu6" [X Link](https://x.com/aditharun_/status/1979189485742231969) 2025-10-17T14:15Z [---] followers, [----] engagements "1/3 Men's testosterone levels did not plummet over the past decade like many wellness podcasts would have you believe. we just switched the measurement assay from immunoassay to mass spectrometry When we switched assays we observed a 100% increase in self-reported healthy men with low testosterone. The threshold needed to be updated when the assay changed. Arun AS Durant TJS El-Khoury JM Krumholz HM. Re-evaluating the threshold for low total testosterone. Clinical Chemistry 2025" [X Link](https://x.com/aditharun_/status/1980357420657393832) 2025-10-20T19:36Z [---] followers, 406.7K engagements "3/3 I found this post where @eryney_ok and @mu_anand rightly pointed out that changes in measurement tools have led people to falsely believe their T is low" [X Link](https://x.com/aditharun_/status/1980357425803956649) 2025-10-20T19:36Z [---] followers, 13K engagements "Druckenmiller offering a masterclass in just over an hour. Will need to re-listen to this again. Takeaways: - Discipline. No matter how good you are once your ego inflates its over - There are many ways to make money in the market the biggest thing is to have a system - Think deeply about where and how you went wrong. People who dont spend a lot of time talking about their failures and focus on their wins are negative signal. - Historical correlations and patterns break when black swan events happen. So standard risk models break during these events and that is exactly when you want risk" [X Link](https://x.com/aditharun_/status/1980412595434918358) 2025-10-20T23:15Z [---] followers, 305.2K engagements "@princetongb $argt" [X Link](https://x.com/aditharun_/status/1980661423324975231) 2025-10-21T15:44Z [----] followers, [---] engagements "@TheIngated I am very interested in figuring out the health effects of microplastics. First places I might look for an association are gut issues and crc in young adults. The hard part is finding microplastic levels in humans indexed with outcomes" [X Link](https://x.com/aditharun_/status/1980819745986478290) 2025-10-22T02:13Z [---] followers, [----] engagements "The entire paper is only [--] page and its attached as an image to the second tweet of my original post. But yes we moved from immunoassay to mass spec and kept using the immunoassay threshold the whole time. The two methods perform largely the same but are not exactly the same. Im not sure that other trace molecules need to have their threshold re-calibrated. If the same story was found for another molecule then yes. But I see it as case by case" [X Link](https://x.com/aditharun_/status/1980823056349061622) 2025-10-22T02:26Z [---] followers, [--] engagements "be a fund manager send a guy to catalogue jackets salespeople at each b2b saas company wore at conferences sales people splurge on jackets when deal flow is up its a discretionary purchase average jacket price for company is a leading indicator of earnings Cheeky work love to see it" [X Link](https://x.com/aditharun_/status/1981554114975637768) 2025-10-24T02:51Z [---] followers, [---] engagements "Interesting hypothesis for anktiva (il15 agonist) in cancer treatment will need to look into data. Immediate thought is that we know il2 boosts nk and t cell counts but it doesnt work well / has many adverse effects Working to explain to the world how and why Bioshield works. T cells and NK cells kill cancer thats irrefutable. Low NK cells and T cells means cancer grows thats irrefutable. So if there is a drug available to boost NK and T cells then there is better chance to kill https://t.co/9nbHmD2coL Working to explain to the world how and why Bioshield works. T cells and NK cells kill" [X Link](https://x.com/aditharun_/status/1981798242808476090) 2025-10-24T19:01Z [---] followers, [---] engagements "@AnilMakam @NEJM @tabuckley_ @arjunmanrai @Gurpreet2015 Agreed What do you think the role of the largely non-procedural physician is in the age of AI My thoughts: Feedback / thoughts appreciated https://open.substack.com/pub/aditharun/p/what-does-a-doctor-do-in-the-ager=4do42&utm_medium=ios https://open.substack.com/pub/aditharun/p/what-does-a-doctor-do-in-the-ager=4do42&utm_medium=ios" [X Link](https://x.com/aditharun_/status/1981818646910345266) 2025-10-24T20:22Z [---] followers, [---] engagements "Will Q4 $TECX Ph1b topline readout of long-acting relaxin (TX45) in pulmonary hypertension (PH) associated with HFrEF be positive Alise (CEO $TECX) said in her Morgan Stanley talk that data is to be expected 2H of October. First by "positive" in this Ph1b I mean two things: 1) its safe in this patient population and 2) it moves the hemodynamic parameters in the right direction. Down the line positive will also mean reduction in symptoms and increased [--] minute walk test but this study measures hemodynamic parameters up to [--] hours after single dose of TX45. So the parameters to judge TX45 in" [X Link](https://x.com/aditharun_/status/1981829320608714986) 2025-10-24T21:05Z [---] followers, 12.8K engagements "@plainyogurt21 ok good to know I heard alise mention it but couldn't find it on azn materials in my search probably just missed something" [X Link](https://x.com/aditharun_/status/1981844325593129251) 2025-10-24T22:04Z [---] followers, [---] engagements "@plainyogurt21 can you clarify what you mean by "if you believe the pathology is actually different where the heart can handle additional CO and the pathology is truly RHD and PH"" [X Link](https://x.com/aditharun_/status/1981844946077524248) 2025-10-24T22:07Z [---] followers, [---] engagements "Ok i think I understand. Lilly looked at acute decompensated hf and tecx/azn is stable hf like you pointed out. And that is the difference imo. Stable means pts have physiologic reserve and thats what counts. Once youre decompensated all bets are off. To me its kind of like saying why doesnt my chemo work in metastatic disease well maybe you should have given it in earlier disease Now your second point on pvr is one i dont agree with. Tecx showz pvr [--] and pvr =2. Pvr declines more when baseline pvr high but that is expected in cpcph since right sided pressures are affected. In ipc pvr is" [X Link](https://x.com/aditharun_/status/1981901383398789368) 2025-10-25T01:51Z [---] followers, [---] engagements "My favorite quote from The Technologic Republic is highly relevant to engineers scientists and physicians: be multidiscplinary. Study things from different fields and apply them. That's what da Vinci Gutenberg and Brunelleschi and others did because they had to" [X Link](https://x.com/aditharun_/status/1982569929674248488) 2025-10-26T22:08Z [---] followers, [---] engagements "AZN ran study in patients with any EF so all comers but if my hypothesis is that it works regardless of EF then we will see an effect. The other relaxin trial papers have supplemental analyses by EF and there was no difference in effect sizes. This is another point that makes me feel lile what happens in hfpef will happen in hfref. Now given the different trial pop in azn and tecx compared to lilly this statement requires a leap of faith because of the different trial pops. I think its a reasonable leap. Yes i do think a chronic stable class II or III HF patient has orders of magnitude more" [X Link](https://x.com/aditharun_/status/1982801885913354303) 2025-10-27T13:29Z [---] followers, [---] engagements "Plain pulmonary vasolidation is no good in hfpef as weve talked about. The relaxin case is gentler vasodilation lusitropy help and with long term use anti-fibrosis - as weve both mentioned. Hard to say if that comes to fruition. I can say that the update in q4 is only hemodynamics not functional improvement so thats all we can care about for this upcoming one Safety is key agreed. Well hopefully be able to get data in the next [--] days" [X Link](https://x.com/aditharun_/status/1982811931464397193) 2025-10-27T14:09Z [---] followers, [---] engagements "Who makes $NTLA's LNP Is it Novartis I believe MoA of LNP is via ApoE on hepatocytes for ATTR programs" [X Link](https://x.com/aditharun_/status/1982826190131065089) 2025-10-27T15:06Z [---] followers, [---] engagements "Recap of $NTLA call Oct [--] 2025: Male patient with ATTR-CM given dose of $NTLA gene therapy on Sep [--] [----]. AST and ALT [--] times upper limit of normal and bilirubin [--] times upper limit of normal. Voluntary pause. Not mandated by FDA. Patient complained of abdominal pain and was evaluated in ED where labs were drawn. AST ALT and bilirubin were elevated. That happened Oct [--] [----]. (*) Clinically the patient is stable per management playing cards with nurse as of last night. Management says "would meet the traditional definition of Hy's Law". Hy's law is a threshold at which fatal drug-induced" [X Link](https://x.com/aditharun_/status/1982833093196050807) 2025-10-27T15:33Z [---] followers, [----] engagements "@JusticeforB_MD SCD Pain crises auto splenectomies and impaired immunity VOCs increasing risk for many things is not solved by current treatment" [X Link](https://x.com/aditharun_/status/1982897809985872122) 2025-10-27T19:50Z [---] followers, [--] engagements "AZN sets the expectations for relaxin in PH thats the bigger trial. If AZN is positive then HFrEF TECX will be positive. If for some reason HFrEF TECX gets announced first then I expect AZN primary endpoint of PVR to hit. Safety in TECX is single dose but AZN is repeated dosing so safety from AZN is a more accurate measure of safety of relaxin. Your theory re: cpcPH is plausible and just to add showed Cpc is the dominant type of PH in HFrEF and HFpEF with like 80% of cases being that way in either etiology of HF. Great discussion really enjoying it https://pubmed.ncbi.nlm.nih.gov/26181215/" [X Link](https://x.com/aditharun_/status/1982947396624519675) 2025-10-27T23:07Z [---] followers, [---] engagements "@mtheory11bio Optimistic they'll release it any day now" [X Link](https://x.com/aditharun_/status/1982972704618586560) 2025-10-28T00:48Z [---] followers, [---] engagements "Lots of talk about Hy's law today after Intellia's Grade [--] adverse event disclosure. Some context into the test characteristics of Hy's law: Hy's Law (ALT or AST 3x ULN & TBili 2x ULN & exclusion of other causes) predicts a 10% risk of death or liver transplantation among affected patients. In a large US managed care cohort Hy's law had a specificity of 92% and a PPV of 2% for acute liver failure among drug-induced liver injury cases. Most patients who get drug-induced liver injury do not progress to acute liver failure. Without the patient's chart its hard to say which way the scales might" [X Link](https://x.com/aditharun_/status/1982974939830403505) 2025-10-28T00:57Z [---] followers, [---] engagements "I do not understand the psychology of the linkedin warrior" [X Link](https://x.com/aditharun_/status/1983002752256286952) 2025-10-28T02:47Z [---] followers, [---] engagements "@3primeAnalytica regn with strong q3 earnings Nice call" [X Link](https://x.com/aditharun_/status/1983149810350993653) 2025-10-28T12:32Z [---] followers, [--] engagements "$VYNE BET inhibitor for nonsegmental vitiligo also canned earlier this year (negative trial results) Another project gone: $INCY's BET inhibitor INCB57643 https://t.co/1HDrjrBmv8 Another project gone: $INCY's BET inhibitor INCB57643 https://t.co/1HDrjrBmv8" [X Link](https://x.com/aditharun_/status/1983150422316470351) 2025-10-28T12:34Z [---] followers, [----] engagements "gene therapies have been derailed by grade 3/4 AEs before and this is a risk to putting money in gene tx companies. Ntla had one asymptomatic patient before which made me think huh thats interesting but the second one soured my appetite. I certainly didnt predict the event and am curious what the official explanation is and how it may have happened biologically" [X Link](https://x.com/aditharun_/status/1983494025316389084) 2025-10-29T11:20Z [---] followers, [--] engagements "Patiently waiting for AZN or TECX updates Alise CEO of TECX said 2H of October which means by Friday EOD Will Q4 $TECX Ph1b topline readout of long-acting relaxin (TX45) in pulmonary hypertension (PH) associated with HFrEF be positive Alise (CEO $TECX) said in her Morgan Stanley talk that data is to be expected 2H of October. First by "positive" in this Ph1b I mean two things: 1) https://t.co/kHLa0Y0Xk4 Will Q4 $TECX Ph1b topline readout of long-acting relaxin (TX45) in pulmonary hypertension (PH) associated with HFrEF be positive Alise (CEO $TECX) said in her Morgan Stanley talk that data is" [X Link](https://x.com/aditharun_/status/1983542018468106468) 2025-10-29T14:30Z [---] followers, [---] engagements "Recently med school classmate (ex-Amazon SWE) and I wrote up some slides on how to think about using AI in medicine with example cases to work through. Prompt engineering evaluating outputs workflow in clinical reasoning etc. Taught it as an interactive session to M2s and will do it for M1s in the coming weeks. It is great to hear and see how they are thinking about and use it - I learn a lot. How we use AI in medicine will continue to evolve rapidly so teaching a specific ways of use may only be valuable for a short time. Instead principles of AI how models are trained and bayesian reasoning" [X Link](https://x.com/aditharun_/status/1983560731871088853) 2025-10-29T15:45Z [---] followers, [---] engagements "The TECX APEX Ph2 trial (topline 2026) and AZN Ph2 trial (expected readout any day now) have the exact same setup except AZN is recruiting PH-HF patients regardless of LVEF and independent of ipcPH/cpcPH designation. TECX seems to be doing a good job working to find the right group of patients that will respond. But by in large if AZN Ph2 succeeds then TECX Ph2 should succeed too. TECX APEX Ph2 trial (topline expected 2026) is evaluating chronic treatment in [---] PH-HFpEF patients enriched for PVR at or above [--] (70% of population) [--] min walk test is secondary endpoint primary endpoints are" [X Link](https://x.com/aditharun_/status/1983913785191534853) 2025-10-30T15:08Z [---] followers, [---] engagements "$REGN Q3 earnings call summary - 3.8B total revenues (up 1% YoY) diluted net income per share of [-----] 1.3B net income. 198M of profit share royalty and licensing. 1.3B R+D expenses (mid-teens increase in expenses in [----] to be expected given trials upcoming). Cash less debt of 16B dollars. - First [--] months of [----] repurchased 2.8B of shares (most for any fiscal year for REGN) anticipate returning 4B in shareholders in dividends and repurchases in [----]. - Dupixent WW net sales 4.9B. 29% YoY growth. 3.6B sales in US 28% YoY growth. Leads market in all established indications. Pulmonogists" [X Link](https://x.com/aditharun_/status/1983931670685208897) 2025-10-30T16:19Z [----] followers, [---] engagements "@3primeAnalytica had a great summary. It is much more succinct than mine and pulled in key quotes from the call would recommend https://x.com/3primeAnalytica/status/1983273052843475443 $REGN what a quarter 🟢12% on earnings & updates: 📍Rev at $3.75B (vs. $3.59B expected +1% YoY). ✅ 📍Non-GAAP EPS of $11.83 exceeded forecasts ($9.64 est.) despite IPR&D charges. GAAP net income rose 9% to $1.46B. ✅ Key drugs: 📍Dupi (what a monster) at $4.86B (+26% https://x.com/3primeAnalytica/status/1983273052843475443 $REGN what a quarter 🟢12% on earnings & updates: 📍Rev at $3.75B (vs. $3.59B expected +1%" [X Link](https://x.com/aditharun_/status/1983932072583479775) 2025-10-30T16:20Z [---] followers, [---] engagements "@drug_smolecules Anticipating AZD3427 announcement Nov [--] Also where did you find the AZN doc you screenshotted in the original post" [X Link](https://x.com/aditharun_/status/1983974443602079796) 2025-10-30T19:09Z [---] followers, [---] engagements "Thanks for sharing Shelby My guess would be that non-platform companies are more likely to succeed because they are focused on only one thing. This finding is quite interesting. How many platform companies actually brought [--] asset to successful phase [--] A lot of public biotech companies bill themselves as platform companies to investors and have a platform "up on their site" but there is maybe [--] or [--] total assets" [X Link](https://x.com/aditharun_/status/1984255686289608830) 2025-10-31T13:46Z [---] followers, [--] engagements "@financebully @HOThomasWPhelps Pbo adjust each metric is my view" [X Link](https://x.com/aditharun_/status/1984274284404252910) 2025-10-31T15:00Z [---] followers, [--] engagements "@HOThomasWPhelps Ok wow nice" [X Link](https://x.com/aditharun_/status/1984283064550523126) 2025-10-31T15:35Z [---] followers, [---] engagements "If curing inherited deafness is not incredible then I don't know what is. A short thread on a beautiful paper 1/3: Otoferlin (synapse transmission of hair cells in ear) deficiency results in congenital deafness. A single intracochlear infusion in one or both ears of a AAV1 gene therapy made of OTO cDNA and hair cell-specific promoter allows these individuals to hear. - [--] of [--] patients can hear [--] dB - [--] of [--] patients treated in both ears achieved [--] decibels. Median response was [--] dB at [--] weeks. - [--] of [--] participants can hear soft speech with gene therapy alone ( [--] decibels) and 3" [X Link](https://x.com/aditharun_/status/1985581914460205345) 2025-11-04T05:36Z [---] followers, [----] engagements "What about the [--] non-responders ( [--] dB at [--] weeks) [--] eventually hit [--] dB at [--] weeks [--] was [--] years old at infusion. Patients [--] [--] and [--] had week [--] levels below [--] dB. However pt [--] eventually did recover. Pt [--] never responded. Pt [--] was [--] years old. Participant [--] (4yo M DB-OTO and implant concomitantly placed) had hearing improvements after [--] weeks but not to primary endpoint level of [--] dB. At week [--] with CI deactivated 100% accuracy at syllable patterns and 50% accuracy for 2-syllable words with no visual cues. Patient [--] the other [--] yo F had a barely above threshold response at week" [X Link](https://x.com/aditharun_/status/1985581917446517010) 2025-11-04T05:36Z [---] followers, [---] engagements "DB-OTO is safe. However note that there is roughly double the AE rate in Part A (one ear) compared to Part B (both ears treated). My guess is that the surgeons and surgical technique get better with more reps. Another hypothesis is that the cochlear implant actually comes with side effects given that [--] of the part A patients concomitantly received implants with gene therapy. [--] of [--] AEs were surgery related. [--] serious AEs: Grade [--] mastoiditis associated with CI in an ear not treated with DB-OTO and Grade [--] walking instability vestibular impairment (recent varicella vaccine). Both resolved" [X Link](https://x.com/aditharun_/status/1985581919900168239) 2025-11-04T05:36Z [---] followers, [---] engagements "Inflation is not 3% Tracking my own basket of goods YoY its more like 10%. I do not think wages are going up 10%. Imagine having dependents.only so far you can stretch a dollar. I remember seeing something about how hamburger helper sales have recently been trending up. not a good sign. - Fed rates are going down - Fed ending QT in Dec liquidity to go up - [--] y T-bill minus [--] y T-bill maturity flat at 0.4% last [--] months constant upward movement since [----]. Growth expected to be slow a bit but not drop too much - US Q2 GDP YoY 2% increase - GOOG R+D 36B https://t.co/XNv1pQfhmt - Fed rates are" [X Link](https://x.com/aditharun_/status/1985769013457113592) 2025-11-04T18:00Z [---] followers, [----] engagements "$TECX down 34% to $16.4/share since Ph1 PH-HFrEF readout. Likely sell the news and then concern over $AZN Ph2 PH-HF readout. Druckenmiller said at some talk I listened to that if the asset is trading down on no news then its a bad thing. I could argue here that there is *news* its just worry over AZN Ph2 and their perceived lateness in announcing results. AZN Q3 earnings 11/6 8a before market opens. The consensus on X is that AZN will announce Ph2 PH study results and/or offer guidance about the program. I'm sure an analysts will ask if not and management's words will be informative (or at" [X Link](https://x.com/aditharun_/status/1985830917664096488) 2025-11-04T22:06Z [----] followers, [----] engagements "$TENX is developing an oral formulation of levosimendan an approved drug in [--] countries (not the US though) for acute decompensated HF (improves symptoms reduce hospitalizations) for PH associated with HFpEF. IV levoseimendan has been tried in PH-HFpEF by TENX actually and the results were reported by Burkhoff et al JACC HF [----]. Notably it improved [--] minute walk test. It didn't move hemodynamics. PO levoseimendan in severe HF did not seem to meaningfully move QoL scores. Trial done by Orion maker of IV levoseimendan. However compared to placebo a net improvement of [--] points in MLHF QoL at" [X Link](https://x.com/aditharun_/status/1986155035986739368) 2025-11-05T19:33Z [----] followers, [----] engagements "Nice large cup of coffee and tuning into astrazeneca's earnings call. good start to the day" [X Link](https://x.com/aditharun_/status/1986417784159400127) 2025-11-06T12:58Z [---] followers, [---] engagements "Doordash a cheeseburger from McDonalds through chase credit card 23% APR Doordash is an IQ test. Do you want a cheeseburger for $27 or will you go pick it up for $6 Doordash is an IQ test. Do you want a cheeseburger for $27 or will you go pick it up for $6" [X Link](https://x.com/aditharun_/status/1986472937495425430) 2025-11-06T16:37Z [---] followers, [---] engagements "For context pluvicto made by Novartis is a radiotherapy that prolongs survival in metastatic prostate cancer. It was first approved in [----] and the label was expanded in [----]. Pluvicto is made of a prostate-specific membrane antigen (PSMA) binder linked to lutetium-177. Once given IV pluvicto binds cells that express PSMA and the lutetium-177 emits beta-minus radiation which induces DNA damage and cell death. This is targeted radioactive destruction of cancer cells by a small molecule similar to radioactive iodine used in treating thyroid cancer. In [----] pluvicto was approved for treating" [X Link](https://x.com/aditharun_/status/1986611274159985148) 2025-11-07T01:46Z [---] followers, [----] engagements "@buccocapital Do chipotle" [X Link](https://x.com/aditharun_/status/1986646580989165636) 2025-11-07T04:07Z [---] followers, [---] engagements "stochastic CRISPR-cas9 indels / frameshifts can lead to either toxic mRNA or an mRNA that is able to break through nonsense mediated decay and create either a toxic or immunogenic protein -- this is because of the specific region of the genome that ntla targets because of the specific crispr machinery ntla uses or just a class-effect in that any frameshift introducer runs this risk or perhaps some combination of the three. clarifying for myself" [X Link](https://x.com/aditharun_/status/1986654307387617505) 2025-11-07T04:37Z [----] followers, [---] engagements "Hacked together an app over the weekend (w a friend from hs) to track hair density from a selfie and applied to YC in [----] and got an email saying top 10% or maybe 5% can't remember It was born from an actual need: my hair was thinning. Basically I got frustrated with being told to try different things (XYZ shampoo oral minoxidil ABC supplement) but there being no actual good way to measure my hair density the metric I actually cared about. If you have an intervention you should have a measurement tool for it beyond "look in the mirror and go off vibes" or "take a photo and manually inspect" [X Link](https://x.com/aditharun_/status/1986661551298748696) 2025-11-07T05:06Z [----] followers, 52.5K engagements "@HairDAO_ Not building custom hardware. Can accept video output from whatever source built initially for standard iphone camera input" [X Link](https://x.com/aditharun_/status/1986835910902591812) 2025-11-07T16:39Z [---] followers, [----] engagements "I got that dog in me The dog in question:" [X Link](https://x.com/aditharun_/status/1986869086349873349) 2025-11-07T18:51Z [---] followers, [---] engagements "@houndcl This was my main concern pre data readout. But bezuclast is less toxic. Wonderful. Hard to predict. What biochemical clues did you see pre-today's data Trying to learn to look at biochem data more" [X Link](https://x.com/aditharun_/status/1987921267995939153) 2025-11-10T16:32Z [---] followers, [---] engagements "@doctorveera Great post What do you make of the bili spike a few days post tx" [X Link](https://x.com/aditharun_/status/1987931748995395943) 2025-11-10T17:13Z [---] followers, [---] engagements "Is there a good long thesis on tractor supply co out there $tsco politicians are buying and i like the company - sales staff in store is great and company is basically a proxy for agriculture industry in US Very reasonable PE. Not knowledgable beyond this" [X Link](https://x.com/aditharun_/status/1987935423423758628) 2025-11-10T17:28Z [----] followers, [---] engagements "@Ridingwaves69 Nobody is paying me can't say if the same is true for others I have no information there. You are long AVXL Please share thesis" [X Link](https://x.com/aditharun_/status/1987964420706770972) 2025-11-10T19:23Z [---] followers, [---] engagements "@3primeAnalytica @Regeneron what do you think about city therapeutics they have ASO for Factor XI advancing to clinic if I remember clinic with timeline of 2029" [X Link](https://x.com/aditharun_/status/1987965126096679322) 2025-11-10T19:26Z [----] followers, [--] engagements "agreed that there is an unmet need agreed that does not cause ARIA as much as current EMA-approved drugs do not understand your point on insolvency we don't treat most patients with AD with anti-amyloid therapy at baseline. it is shared decision making and certainly not stadnard of care. i also disagree i think most docs treating AD patients may not use once they read trial but I don't hold this point strongly given that it could change placebo is also safe . avxl doesn't beat the current donanemab results.and no mechanistic plausability unlike donanemab imo.also lots of weird stats /" [X Link](https://x.com/aditharun_/status/1987972315368571073) 2025-11-10T19:55Z [---] followers, [--] engagements "@adamfeuerstein If creating a bell labs was as easy as getting some cash together and hanging a sign that says bell labs we would have solved quantum computing by now Capital with no strings attached is a great start but like arena there are strings attached especially when softbank comes in" [X Link](https://x.com/aditharun_/status/1988044347640582403) 2025-11-11T00:41Z [---] followers, [---] engagements "I don't think you can create a Bell Labs just because you raised some money from sama and softbank. At some point (sooner rather than later) softbank is going to expect return. And research especially at the cutting edge is incredibly uncertain - discoveries are not churned out every quarter. Bells Labs worked because there was a company with a monopoly over an industry that funded its work without strings attached. After bell was broken up funding dried. softbank can only commit so much capital even if it is a couple billion or some crazy number. It's preferable that the lab be attached to a" [X Link](https://x.com/aditharun_/status/1988052670653891063) 2025-11-11T01:14Z [---] followers, [----] engagements "@Drlipid @Amgen Woah did not put [--] and [--] together. Thats pretty cool naming from them" [X Link](https://x.com/aditharun_/status/1988056666047152510) 2025-11-11T01:30Z [----] followers, [---] engagements "@Sanjusinha7 Just my vibes in terms of rank prioritizing targets. Is this not consensus in preclin drug dev To argue against myself chemo is broadly toxic and they are standard in many fields" [X Link](https://x.com/aditharun_/status/1988295651038556403) 2025-11-11T17:19Z [---] followers, [--] engagements "If true $nvo evoke will read positive We will know Dec [--]. Im not sure why but novo management said they didnt know or didnt want to know cant remember (someone please correct) Apparently the connection between GLP-1 and AD is not so far fetched. Cleveland Clinic did a real-world observation study and found that patients who initiated GLP-1s had 31% lower risk of developing AD. SGLT2i had 33% reduction. DPP4i was the control. https://t.co/xWOyJTWNkk https://t.co/FBMHGpAipI Apparently the connection between GLP-1 and AD is not so far fetched. Cleveland Clinic did a real-world observation study" [X Link](https://x.com/aditharun_/status/1988446778140581993) 2025-11-12T03:20Z [---] followers, [---] engagements "@avidresearch If reads positive big market unlock and maybe first real tx for AD" [X Link](https://x.com/aditharun_/status/1988455674112819281) 2025-11-12T03:55Z [---] followers, [--] engagements "@avidresearch Not for nothing but iirc burry disclosed 2% of portfolio (notional value) in pfizer calls Pfizer is a winner bc of metsera deal if novo reads out positive on evoke @avidresearch brought this up earlier today" [X Link](https://x.com/aditharun_/status/1988456110227918901) 2025-11-12T03:57Z [---] followers, [--] engagements "If true $nvo evoke will read positive We will know Dec [--]. Im not sure why but novo management said they didnt know or didnt want to know cant remember (someone please correct) Apparently the connection between GLP-1 and AD is not so far fetched. Cleveland Clinic did a real-world observation study and found that patients who initiated GLP-1s had 31% lower risk of developing AD. SGLT2i had 33% reduction. DPP4i was the control. https://t.co/xWOyJTWNkk https://t.co/FBMHGpAipI Apparently the connection between GLP-1 and AD is not so far fetched. Cleveland Clinic did a real-world observation study" [X Link](https://x.com/aditharun_/status/1988456224040370188) 2025-11-12T03:58Z [---] followers, [---] engagements "@therealRYC @NatureMedicine Theyre basically the same imo" [X Link](https://x.com/aditharun_/status/1988816940417298663) 2025-11-13T03:51Z [----] followers, [---] engagements "@therealRYC @NatureMedicine Interesting that TZP has accelerating market share and semaglutide is decelerating TZP has rate of weight loss but you basically end up in the same place wirh a long enough time horizon which is what matters for non-aesthetic reasons" [X Link](https://x.com/aditharun_/status/1988817404877787616) 2025-11-13T03:53Z [----] followers, [---] engagements "$MRSN up 210% $DAWN to acquire for 25/share with cash up to 30.25/share 🤯 tbh i didnt have this on my radar There is no good therapy for patients with metastatic TNBC who have many lines of prior therapy including at least [--] prior topo-1 ADC (like sacituzumab govitecan) median PFS is 1-3 months and median OS is 6-8 months. If patients are HER2 low (so not truly TNBC) they may get https://t.co/N0otXrfHby There is no good therapy for patients with metastatic TNBC who have many lines of prior therapy including at least [--] prior topo-1 ADC (like sacituzumab govitecan) median PFS is 1-3 months and" [X Link](https://x.com/aditharun_/status/1988977654306107584) 2025-11-13T14:30Z [----] followers, [----] engagements "@varma_ashwin97 @therealRYC @NatureMedicine @therealRYC and I reporting for duty @novonordisk" [X Link](https://x.com/aditharun_/status/1988996702192165354) 2025-11-13T15:45Z [----] followers, [--] engagements "I am surprised (and confused) how $avxl is still at a 320M market cap this morning" [X Link](https://x.com/aditharun_/status/1989343878550524127) 2025-11-14T14:45Z [----] followers, [----] engagements "@Banana_Oncology @jimcramer Inverse cramer is going to be wonderful for xbi" [X Link](https://x.com/aditharun_/status/1989445482536718476) 2025-11-14T21:28Z [----] followers, [---] engagements "CEOs of the biggest companies on the planet are shitposting at an unprecedented rate We have entered a new world (See sundar pichai posting about polymarket odds for his own products next release) Soon. https://t.co/KoEA6KjnO8 Soon. https://t.co/KoEA6KjnO8" [X Link](https://x.com/aditharun_/status/1990136933285081178) 2025-11-16T19:16Z [----] followers, [----] engagements "Forecasting $NVO's EVOKE/EVOKE+ semaglutide in early-stage Alzheimer Disease trial results Open to discussion please share thoughts and comments. https://www.aditharun.com/p/semaglutide-in-early-stage-alzheimers https://www.aditharun.com/p/semaglutide-in-early-stage-alzheimers" [X Link](https://x.com/aditharun_/status/1990421163588092388) 2025-11-17T14:06Z [----] followers, 25.2K engagements "@blippd Sentiment is very low and AD trial takes many years. But how many patients can they actually capture As in patients who dont already rule in for MACE or T2DM or other indications. I think 20% may happen if readout is above that seen in anti amyloid ab drug performance" [X Link](https://x.com/aditharun_/status/1990476694092984579) 2025-11-17T17:46Z [----] followers, [---] engagements "Agreed Safety is key and we know glp1 side effect profile well this trial wont unlock new safety signals. At that point its up to patients and doctor shared decision making imo. I favor this especially because there is a high unmet need and we have no good option so semaglutide is better than nothing or incredibly expensive brain bleed risk anti amyloid drugs" [X Link](https://x.com/aditharun_/status/1990536006320136353) 2025-11-17T21:42Z [----] followers, [---] engagements "PH-HFpEF space heating up 🔥 - $tecx relaxin mimetic ph2 [----] - $tenx levosimendan (inodilator) ph3 [----] - $azn relaxin mimetic ph2 readout [----] (still waiting) - $mrk activin inhibitor ph2 positive (announced today) / already approved in PAH (group [--] PH) to initiate phase [--] There is a large unmet need for patients with PH including group [--] PH (PH associated with left heart disease). It is wonderful that by end of 2026/2027 we could have new treatment options for this disease Will Q4 $TECX Ph1b topline readout of long-acting relaxin (TX45) in pulmonary hypertension (PH) associated with HFrEF" [X Link](https://x.com/aditharun_/status/1990824935451758811) 2025-11-18T16:50Z [----] followers, 21.9K engagements "@ResearchPulse1 What can I read on manufacturing side to learn about small molecule and peptide synthesis costs logistics and procedure" [X Link](https://x.com/aditharun_/status/1990828358758768722) 2025-11-18T17:04Z [----] followers, [--] engagements "@ResearchPulse1 @researchitgg Oh I see. But who cares Adverse effects matter but so far its pretty similar. And the 73x only matters insofar as it influences cost of the whole system (making and distributing pill vs injectable)" [X Link](https://x.com/aditharun_/status/1990840442276004284) 2025-11-18T17:52Z [----] followers, [--] engagements "@ResearchPulse1 @researchitgg Oral small molecules scale faster. But oral glp1 peptide vs sq injection is the question and lly/nvo are ramping oral production and pill stockpiles. Distribution is easier too" [X Link](https://x.com/aditharun_/status/1990850335234011153) 2025-11-18T18:31Z [----] followers, [--] engagements "@avidresearch @DrSamuelBHume AZN still hasn't shared data but also hasn't killed their program afaik TECX and TENX no expected readouts till [----] So all is quiet till end of Q1 [----] except for AZN" [X Link](https://x.com/aditharun_/status/1990928778780983577) 2025-11-18T23:43Z [----] followers, [---] engagements "@plainyogurt21 $MRK is going after a population with more severe disease and RV dysfunction (if you accept that PVR is a good marker of RV dysfunction in HFpEF)" [X Link](https://x.com/aditharun_/status/1990936009387720949) 2025-11-19T00:11Z [----] followers, [---] engagements "In other words $fulc (600M mc) is trading at a little under 1/3 the value of $beam (2.1B mc) $FULC +18% to 52w high off $AGIO p3 P1b 20mg cohort data at ASH $FULC +18% to 52w high off $AGIO p3 P1b 20mg cohort data at ASH" [X Link](https://x.com/aditharun_/status/1991191488131121451) 2025-11-19T17:06Z [----] followers, [----] engagements "Who will generate successful phase 2/3 data for AI-designed antibody first Assume bindcraft coerced to make antibody designs only Nabla Chai Other / show me results BindCraft Nabla Chai Other / show me results BindCraft" [X Link](https://x.com/aditharun_/status/1991541975057854858) 2025-11-20T16:19Z [----] followers, [---] engagements "@3rdFloorCapital Yeah. Traded up AH and early session but now down. Maybe this is just mild rerating given jobs report and now there will be no cuts Prediction markets had 60% chance of rate cut a few days ago now its like 20% so maybe this is a function of that news idk" [X Link](https://x.com/aditharun_/status/1991565584098832525) 2025-11-20T17:53Z [----] followers, [---] engagements "$FULC is the most promising sickle cell disease drug currently on the market. It also has the least data available. I will articulate my reasoning here: Gene editing approaches require chemo conditioning requiring in patient admission dealing with chemo toxicities (pretty bad) high up front cost (millions of dollars). There is an ongoing question as to whether the median patient wants this. Scalability remains a concern. $CRSP and $VRTX make Casgevy (FDA approved) -- apheresis busulfan conditioning then infuse ex-vivo edited hematopoietic stem cells (cas9 double stranded break of bcl11a" [X Link](https://x.com/aditharun_/status/1991611890213220726) 2025-11-20T20:57Z [----] followers, 16.4K engagements "Lumina is an interesting concept but I don't think the scientific evidence for its efficacy exists. Also we need human testing data. I get not wanting to register and go down FDA path but some human testing data is warranted before selling. Thoughts below but TL;DR I think you actually create periodontal issues with the probiotic. Also caries scoring system is based on [--] years ago and vague/not validated/highly subjective. Lumina probiotics per their company website is a probiotic optimized through biodesign to rebalance your oral pH and inhibit the growth of acid-secreting bacteria to" [X Link](https://x.com/aditharun_/status/1991964784875762019) 2025-11-21T20:19Z [----] followers, [----] engagements "@gen0m1cs I see it the same way. Downside risk is real given the scientific backing this product stands on. And I suspect this is something where cumulative exposure will matter. This also makes it harder to test because we would need to observe outcomes over a longer period of time" [X Link](https://x.com/aditharun_/status/1991979458916724756) 2025-11-21T21:18Z [----] followers, [--] engagements "@princetongb Whats your abvx pt for buyout" [X Link](https://x.com/aditharun_/status/1992671544883958176) 2025-11-23T19:08Z [----] followers, [---] engagements "I was wrong semaglutide does not change clinical dementia rating score trajectories meaningfully compared to placebo in patients with mild dementia due to Alzheimer's disease. This is a setback for patients but something that I felt needed to be tested given the reasons I covered in the quoted post. It may be that we did not offer the intervention early enough in the disease course. By the time patients with Alzheimer's disease are symptomatic / have imaging findings the disease process has been in motion for at least a decade. Perhaps an intervention like GLP1s that dampen inflammation and" [X Link](https://x.com/aditharun_/status/1993003669390807521) 2025-11-24T17:07Z [----] followers, 14.4K engagements "From the $nvo stock perspective I anticipated a plus/minus 5-10% move on trial results as article in my substack post a week ago I leaned towards the trial hitting statistical significance in CDR-SB but was not overwhelmingly sure. Net-net I felt the risk-reward was there given the anticipated move and the way the market was thinking about this trial readout. Sized my position accordingly and was wrong. $NVO fell 10% immediately after announcment and is now down 6% - within my expectations. Totally fine with the result and happy with sizing and way I structured trade. http://www.aditharun.com" [X Link](https://x.com/aditharun_/status/1993006806394683896) 2025-11-24T17:20Z [----] followers, [----] engagements "The stage of disease matters and in the re-analysis of the RCTs of GLP1-RAs time to dementia was the endpoint that showed benefit [--]. Incidence of disease is different from progression of disease. GLP1 for risk factor control and anti-inflammation WAY before disease even manifests clinically may be very useful. Perhaps individuals who are homozygous for APOE4 may be poised to receive benefit from VERY early use of GLP1s [--] Nrgaard et al. Treatment with glucagon-like peptide-1 receptor agonists and incidence of dementia: Data from pooled double-blind randomized controlled trials and nationwide" [X Link](https://x.com/aditharun_/status/1993007558232080451) 2025-11-24T17:23Z [----] followers, [---] engagements Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing
@aditharun_ Adith Arun
Adith Arun posts on X about $pasg, $tecx, $rvmd, in the the most. They currently have [-----] followers and [---] posts still getting attention that total [------] engagements in the last [--] hours.
Engagements: [------] #
- [--] Week [------] +28%
- [--] Month [------] +8.70%
- [--] Months [-------] +2,185%
Mentions: [--] #
- [--] Week [--] -26%
- [--] Month [--] +2.40%
- [--] Months [---] +261%
Followers: [-----] #
- [--] Week [-----] +1.20%
- [--] Month [-----] +3.10%
- [--] Months [-----] +355%
CreatorRank: [---------] #
Social Influence
Social category influence stocks finance currencies cryptocurrencies technology brands countries social networks
Social topic influence $pasg, $tecx #5, $rvmd, in the, $prax, $imrx, $fulc, $rzlt, $lly, has been
Top accounts mentioned or mentioned by @plainyogurt21 @3primeanalytica @houndcl @avidresearch @capitalshipyard @princetongb @therealryc @bananaoncology @adamfeuerstein @drsamuelbhume @zhaoweiasu @financebully @naturemedicine @researchpulse1 @medstudentinvst @rohitsjhawar @applehelix @emcclintock @bermudadude @biotech2k1
Top assets mentioned Passage (PASG) Revolution Medicines, Inc. (RVMD) Praxis Precision Medicines, Inc. (PRAX) Immuneering Corporation (IMRX) Eli Lilly and Company (LLY) AstraZeneca PLC (AZN) Novo-Nordisk (NVO) uniQure N.V. (QURE) Morgan Stanley (MS) BEAM (BEAM) Prime Medicine, Inc. (PRME) Elicio Therapeutics, Inc. (ELTX) Regeneron Pharmaceuticals Inc (REGN) Arvinas, Inc (ARVN) PMV Pharmaceuticals, Inc. (PMVP) BridgeBio Pharma, Inc. Common Stock (BBIO) Alphabet Inc Class A (GOOGL) Sarepta Therapeutics, Inc. (SRPT) Viking Holdings Ltd (VIK) Xenon Pharmaceuticals Inc (XENE) Novartis AG (NVS) Merck & Co., Inc. (MRK) Fractyl Health, Inc. (GUTS)
Top Social Posts
Top posts by engagements in the last [--] hours
"Does the threshold for low total testosterone need to be re-evaluated New in @Clin_Chem_ADLM @ThomasDurantMD @ClinChemJoe @hmkyale and I use NHANES data to suggest that the traditional [---] ng/dL threshold may need to be revisited 🧵 1/7"
X Link 2025-05-05T10:17Z [---] followers, [--] engagements
"Designing antibody drugs depends on developability and binding both now capable of being modeled. Shehata et al [----] [--] report the binding-developability tradeoff which can be incorporated into a global loss function where developability and binding are modeled independently Great new resource for anyone working on antibody CMC developability or discovery. New preprint out today from Ginkgo that benchmarks [---] therapeutic IgGs across [--] developability assays. The resulting dataset is ML-ready and publicly available. If you're thinking about early https://t.co/C9xuHYZOu8 Great new resource for"
X Link 2025-05-06T15:33Z [---] followers, [---] engagements
"Approved mAbs have melting temps on par with Abs with the highest level of mutations (i.e highest binding affinity). Both have melting temps [--] degrees lower than native Abs so are less stable than native Abs. This suggests that we should be more willing to accept lower developability if it improves binding given we know that approved mAbs operate at this level. Our global loss function should reflect this"
X Link 2025-05-06T15:33Z [---] followers, [--] engagements
"Improving the individual models for predicting binding and developability along with playing with ensembling method are key. I love the recent open sourcing by @Ginkgo here and companies like @tahoe_ai - itll help progress happen quicker"
X Link 2025-05-06T15:33Z [---] followers, [---] engagements
"One caveat: Binding affinity is indirectly measured by # of mutations in [--] an imperfect measure (Km would be better) so Im uncertain of the effect size of the developability-affinity tradeoff"
X Link 2025-05-06T15:33Z [---] followers, [--] engagements
"What model underlies FDAs process here for genAI in these tasks presumably openAIs models given their recent meetings How are they verifying accuracy with their aggressive rollout plan and monitoring for model drift and hallucination Theres not much information on this afaik from jeremy walshs office or makary"
X Link 2025-05-10T16:23Z [---] followers, [--] engagements
"I think gnomAD puts pathogenic ttr variants at 57ish per 100k so in the world probably 5M ppl w the variants of interest. Assume majority are asx or minimally sx. Variants are enriched in those of African ancestry northern portugal taiwan japan so start w big hospitals in these regions. Assume 1% conservatively have the variant and are asymptomatic in an enriched region. So need to sequence 60k ppl. Cost of targeted sequence assay to detect is [--] usd () so cost is $3M if all goes well"
X Link 2025-05-13T12:42Z [---] followers, [--] engagements
"Colibactin signature was shown to be positive in 50% of CRC patients younger than [--]. Colibactin signature positivity steadily decreased as age of onset of CRC increased. Only 15% of 70+ year olds were colibactin positive [--]. Why And what does this mean Some thoughts: (1/)"
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"Should we care Yes. Colibactin signatures are enriched in early clonal compared to late clonal mutations similar to other well known clock-like signatures that may promote carcinogenesis. Evidence suggests colibactin induces mutagenesis early in life (within the first decade) possibly by APC inactivation [--]. In [--] they presented data that colibactin signatures accounted for a sizeable fraction of APC indel drivers and preferentially did so compared to other genes like p53 which occur later. (2/)"
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"If we believe colibactin can inactivate APC we can take pts w positive colibactin signature in their cancer tissue and look at their normal colon tissue. We should expect to see normal crc tissue also have colibactin positive signatures associated with APC changes. Do their normal tissues show this signature If true this would offer more strength to the colibactin idea. (3/)"
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"Could it be early exposure to antibiotics causing proliferation of colibactin bacteria C-section Probiotic use with bacteria capable of making colibactin Microplastics Low fiber diets (5/)"
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"Low fiber damages colon lining. The association with colibactin positive crc with western diet (low fiber) has been shown [--]. Colibactin producing bacteria thrive in low fiber diets and opportunistically take residence upon damage. With a shift in standard eating habits people may unintentionally consume a low fiber diet increasing odds for CRC. How does low fiber mediate tumorigenesis Is it due to a general inflammatory state or does it allow for some other process like colibactin producing bacteria to act (6/)"
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"Other signatures reported in [--] like sbs89 dbs8 and sbs94 were observed preferentially in argentinian and colombian patients. Where are these signatures coming from Argentineans tend to eat more red meat which we know increases odds for crc so maybe these signatures reflect a diet exposure (e.g. red meat low fiber). (7/)"
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"Microplastics could damage the colon too leading to pro-tumor state. This has long been a hypothesis in the field. The hard part is nearly everyone in the last [--] decades has been exposed to microplastics making it hard to parse effect. We would also expect a large delay in time between exposure and effect like smoking and lung cancer. But if causal there should be a time and dose relationship (highest microplastic level and longest time under microplastic influence from birth). imagine area under time-dose curve and look at highest quartile of these individuals might offer the largest signal."
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"The microplastics CRC question reminds me of the lung cancer smoking connection where the majority of people smoked making it hard to figure things out. Then doll and hill came with their cohort study asking physicians if they smoked then watched the death reports to show the strong relationship. We can do this for microplastics too but we will first need to define microplastic exposure. (10/)"
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"I am very interested in trying to understand the reason behind the uptick in CRC in young people. This phenomenon has been highlighted by many including the annual @AmericanCancer report @VinGuptaMD @theNCI and many others. (11/)"
X Link 2025-05-16T20:01Z [---] followers, [--] engagements
"I agree that risk group is undefined and in an undifferentiated group of patients nnt would be hard to estimate. Being as generous as we can and saying nnt of [--] (based on trials) with a generic statin 300k seems really high. I would have to assume the pcsk9i increase the cost Either way lots of questions with this data. Maybe they count cholesterol screening cost is so low because the lab assay is so cheap to run and then only the exceedingly high values that could lead to pt dying in a year are counted"
X Link 2025-05-17T19:22Z [---] followers, [--] engagements
"@AdamRodmanMD @MedEdFlamingo What company does the llm detection on multiple choice tests"
X Link 2025-05-31T23:06Z [---] followers, [--] engagements
"@kenbwork Patients on this tx are going to need to be on immunosupression probably lifelong. Comes with a number of risks. CGM + auto-infusers of insulin to keep A1C and blood glucose titrated seem like better options for patients"
X Link 2025-06-22T20:59Z [---] followers, [---] engagements
"@Biotech2k1 New to public equity biotech investing do you mean biotech stocks will crash along w rest of market (qqq ath) or separately from rest of market"
X Link 2025-06-25T17:34Z [---] followers, [---] engagements
"@Biotech2k1 @TohJah7809 Why Just because it has good binding affinity for [--] things so their tox are additive"
X Link 2025-07-03T15:18Z [---] followers, [--] engagements
"I want to RL some bio stuff and have a few specific problems in mind. What do I read to get up to speed on the RL stuff How do I know if my data has enough inherent information in it to be RL-able Biology is very messy. @willccbb @khoomeik and others DMs welcome"
X Link 2025-07-08T19:05Z [---] followers, [----] engagements
"Yeah good q. Verifier is straightforward - I have a binary yes/no flag for each sample (so I just need to map the LLM output to 0/1 space maybe cross-entropy). Task is to predict the probability of the outcome happening. The input data for each sample contains 1.n many rows with each row representing a certain point in time before the event happens (some samples may have [--] rows some may have [--] and some may have n in unevenly spaced timepoints). The times across sample are irrelevant. For each time-point for each sample I have p many structured data fields. From this nxp matrix for each sample"
X Link 2025-07-08T21:14Z [---] followers, [---] engagements
"Verifier is straightforward - I have a binary yes/no flag for each sample (so I just need to map the LLM output to 0/1 space maybe cross-entropy). Task is to predict the probability of the outcome happening. The input data for each sample contains 1.n many rows with each row representing a certain point in time before the event happens (some samples may have [--] rows some may have [--] and some may have n in unevenly spaced timepoints). The times across sample are irrelevant. For each time-point for each sample I have p many structured data fields. From this nxp matrix for each sample I wish to"
X Link 2025-07-08T21:15Z [---] followers, [---] engagements
"@houndcl I agree that w ur assessment on crdf readout outcome. Naked short is a lot of risk. Any other ways to be synthetically short while keeping risk down given stocks like crdf dont have options markets"
X Link 2025-07-09T19:29Z [---] followers, [---] engagements
"@MichaelOkun What would the net harm be of giving patients with PD or at high risk for developing it nicotine patches Addiction risk is low and may confer benefit. RCT before doing or collecting more RWE is needed but curious to hear your thoughts"
X Link 2025-07-14T12:31Z [---] followers, [--] engagements
"The majority (60%) of genetic diseases in hospitalized infants are single nucleotide substitutions. Base editing ($BEAM) as accomplished by Kiran Musunuru and team for Baby KJ and prime editing ($PRME) can fix these. We have to 1) sequence infants with suspected genetic disorders and figure out IF the variant is actually harmful 2) platform N-of-1 gene editing treatments Of the [---] variants identified in infants with a suspected genetic disorder by either WGS or targeted sequencing [---] were single nucleotide substitutions in a [----] JAMA study. The hard part is figuring out if a variant is"
X Link 2025-07-15T21:24Z [---] followers, [---] engagements
"@SinaHartung Have you tried using the open evidence API to build Eureka"
X Link 2025-07-18T15:05Z [---] followers, [--] engagements
"@nikillinit The company that serves the LLM needs a ad marketplace where others can bid and buy mentions similar in spirit to current ad search methods. That said google is probably in the best place to do this well"
X Link 2025-07-21T16:02Z [---] followers, [--] engagements
"How can we accurately classify cause of death in death certificates Data on US mortality rates and trends for specific causes of death including my own work comes from the cause(s) of death listed on the death certificate. And these reports are largely filled out by overworked residents and doctors who are trying to figure out what happened to a patient who died out of the hospital in between notes/rounding/clinic/other duties. This is what I see in the hospital and @adamcifu had the same experience [--]. Take for example this paper [--] showing 51% of death certificates had major errors. Expert"
X Link 2025-07-21T18:11Z [---] followers, [---] engagements
"* The current medical system does not compensate physician time to adjudicate cause of death. It is important work though. This is a great opportunity for an LLM to step in. AI can read the chart offer its guess as to cause of death as a second data point on the patient's cause of death. Researchers can take this extra data point into account in their work"
X Link 2025-07-21T18:11Z [---] followers, [--] engagements
"@dylan522p Is this Taiwan equivalent ISEF or a regular high school exhibition Would be wild if latter"
X Link 2025-07-22T19:50Z [---] followers, [----] engagements
"$srpt's viral vector appears to trigger liver failure in a small subset of patients that led to [--] deaths. And the 3rd death in the limb girdle is probably related. What immunosuppression regimen can $srpt use to prevent immune cascade Do we have to immunosupress everyone What supporting therapies can we use to prevent acute liver failure if even mild liver enzyme spikes are observed How can we identify which minority of patients will experience serious liver injury The viral vector is the common thread here. It is reasonable that FDA revoked platform designation for $srpt viral vector in"
X Link 2025-07-23T00:46Z [---] followers, [---] engagements
"* From what I gather: 2d from infusion: viral capsid is presented on hepatocyte MHC1 and viral dna sensed by innate immune system (liver is main sink for $srpt viral vector). Innate immune system triggers cytokine release and local immune reaction causing small vessel thrombosis / angiopathy. few weeks: T cells against the capsid proliferate and return to liver 1-2 months: T cells attack capsid presenting hepatocytes (liver enzymes flare) 2-3 months: depending on extent of damage from thrombosis and hepatocyte death patients can progress to liver injury"
X Link 2025-07-23T00:46Z [---] followers, [---] engagements
"@john_hersc79276 Agreed that fighting FDA is not a good strategy. The goal is to work together"
X Link 2025-07-23T00:47Z [---] followers, [--] engagements
"@princetongb Wow. compared to viking having [--] fte"
X Link 2025-07-24T16:06Z [---] followers, [--] engagements
"$CRDF observed in a post-hoc Ph2 analysis that bevacizumab-naive KRAS mutant mCRC patients (n=13) treated with bevacizumab+FOLFIRI+onvansertib had a median PFS of [----] and ORR of 77%. They state that the "exceptional efficacy observed in our study in bevacizumab-nave patients is unique to onvansertib". And later the post-hoc analysis "underscores the unique sensitivity of this population to onvansertib. This exceptional response alongside our translational studies showing onvansertibs effect on the hypoxia pathway and angiogenesis supports shifting onvansertib development to the frontline"
X Link 2025-07-24T21:04Z [---] followers, [----] engagements
"Nothing rubs me the wrong way more than a friend venmo requesting $4.18 for my share of the six pack he bought"
X Link 2025-07-26T14:36Z [---] followers, [--] engagements
"The technology underlying companion diagnostics determine who can receive the drug. It dictates who can even have a chance at realizing benefit from the drug. The major inclusion criteria for adagrasib is a KRAS G12C mutation but executing to detect this mutation can leave patients out. Here are some ways: WGS at 80x coverage of tumor identified KRAS G12C mutations in [--] of [--] lung adenocarcinoma patients previously labelled as RAS/RTK/RAF negative from WES [--]. These patients would have previously not been eligible for adagrasib/sotorasib but could have benefited if sequencing had identified"
X Link 2025-07-29T16:32Z [---] followers, [---] engagements
"The companion diagnostic for adagrasib [--] uses a patient's cfDNA sequenced on Illumina NovaSeq6000 to detect KRAS G12C mutations if the log likelihood score is above a threshold. The log likelihood score is based on mutant allele fraction (variant reads / total reads) and unique variant molecules to reflect the likelihood of the variant being true signal over noise. If we use MAF from DNA isolated from WBCs or a matched normal we could determine if the cfDNA variant is background or not with more sensitivity"
X Link 2025-07-29T16:32Z [---] followers, [--] engagements
"@Biotenic I only saw [--] reddit post couldnt find any others which made it tough to assess"
X Link 2025-07-31T18:44Z [---] followers, [--] engagements
"A [--] nucleotide deletion (delGT) in NCF1 which encodes p47 a subunit of NADPH oxidase renders NADPH oxidase defective. This prevents neutrophils from generating reactive oxygen species. Patients get recurrent bacterial and fungal infections typically catalase-positive bugs. Stem cell transplant from a donor is the only potential cure. Enter prime editing. $PRME uses prime editing to correct the delGT and restore neutrophil NADPH oxidase activity. It's an ex vivo solution similar to Casgevy - a patient's stem cells are collected the prime editing system is added and the edited cells are"
X Link 2025-07-31T18:51Z [---] followers, [---] engagements
"Beam licenses Acuitas LNPs the same class of LNPs Verve uses. Prime uses their own GalNAc-LNP. LNPs can have significant safety events so will be interesting to see how Prime's own LNP performs. Also what type of LNP will Prime use for their CF program SORT LNPs are a good choice imo. Lipid chemistry is fascinating"
X Link 2025-07-31T21:16Z [---] followers, [---] engagements
"@alamentarius How does mfn change some of this Not sure how to think about pricing"
X Link 2025-07-31T22:36Z [---] followers, [---] engagements
"@BalaBioResearch Perhaps market feels execution of sales insurance coverage or executive team operating capability is doubted. Not sure something to investigate any thoughts"
X Link 2025-08-02T23:55Z [---] followers, [---] engagements
"The biology is interesting. TL;DR instead of fixing the issue in AGXT they try to prevent the precursor of the toxic compound from being created by introducing more mutations into a different gene (HOA1). The substrate reduction approach. Primary hyperoxaluria Type [--] (PH1) comes from mutations in AGXT that turns off alanine-glyoxylate aminotransferase activity. This forces glyoxylate to buildup and be converted into oxalate. Oxalate accumulation causes symptoms (kidney stones growth delay bone pain anemia cardiac issues non-healing ulcers). Arbor approaches PH1 not by trying to fix AGXT but"
X Link 2025-08-03T15:50Z [---] followers, [---] engagements
"Best-in-disease may be a little premature but $PRAX vormatrigine results are good very comparable to xenon. Patients w [--] seizures on average [--] anti-seizure meds at baseline and vormatrigine has safety profile on par with other anti-epileptics (worse than xenon) + 56% median reduction in seizure burden at week [--] + 22% are at zero burden during observational period. Safety is a little bit worrying compared to xenon data but otherwise great results"
X Link 2025-08-04T13:46Z [---] followers, [----] engagements
"Notes from $PRAX call this AM: Vormatrigine d/c rate due to improper titration of background ASMs. Going to be more deliberate in dose reduction of background ASMs No food effect w vormat. Xenon has with food label. Xenon has once daily oral dosing w no titration like vormat. But half life much longer (24 vs 6ish). Most patients are women. Xenon and prax drugs so far not noted to cause problems with contraception but long term data needed. Prax has no cyp induction but xenon asks that barrier method be used if patient on hormonal only. Potential advantage praxis here. Adding [--] mg vormat arm"
X Link 2025-08-04T21:40Z [---] followers, [---] engagements
"$XENE excluded cenobamate in Phase 2b and [--]. $PRAX does not shy away in cenobamate group (iirc 30% of patients on hefty doses) 55% median reduction in seizure burden. I would be very interested in seeing the AE stratified into cenobamate users versus not"
X Link 2025-08-04T21:55Z [---] followers, [---] engagements
"Power-1 readout w placebo will be key for evaluating vormatrigine. August conference is going to be more $prax radiant data I think. That said I see no reason the science will not work. AEs and deepening of responses are two of the most salient concerns. AEs would be great if prax did subset by cenobamate because they kept saying that on the call w no data shown to back it up. If you believe this + background asm titration being mismanaged then its fine. Im not a neuro/psych doc so would love some input here. Xenon didnt report the same concern w background asm but they also didnt have"
X Link 2025-08-06T01:45Z [---] followers, [---] engagements
"PURE speculation: Head of hypoimmune program at $SANA (Sonja) left to work on research and questions she is really curious about (and has always been). Perhaps she is not particularly interested in seeing the rest of the clinical studies / execution (compared to opportunity cost of working on hypoimmune organs and other research questions) if a lot of it is "solved" and many other MDs have the knowledge/experience to see it through well. She started as a Professor at Cedars-Sinai in Aug [----] along w appointment at SANA per LinkedIn. So perhaps she agreed/wanted to stay on till the initial"
X Link 2025-08-07T00:05Z [---] followers, [----] engagements
"Borzoi is also directionally correct at identifying expression changes of a SNP. AUROC 0.8+ and AUPRC 0.9+ when SNP is within 3kb of TSS. However it does not do a good job at estimating effect size of SNP (see panel d) w/ Spearman correlation of [---] and doesn't tell you how rare a variant is (as we expect from modeling setup)"
X Link 2025-08-07T17:21Z [---] followers, [---] engagements
"RFDiffusion can make a high-affinity (pretty) specific de novo binder from just the target protein sequence. Feed in sequence of protein you want to bind to -- RFDiffusion makes designs -- Protein MPNN fleshes out sequence of binder -- AF3 folds and filters based on how plausible the pose is (you can combine with binding affinity predictors too) -- experimentally synthesize and test binding affinity. This goes beyond the functional motif scaffolding problem where you know a binding motif and want to optimize that in a future design. RFDiffusion does that too but the hard part is knowing a"
X Link 2025-08-08T12:44Z [---] followers, [----] engagements
"More than twice the number of Oreos sold in [----] are sold this year despite 12% of Americans having tried a GLP-1 (6% currently on it). [--] billion Oreos were sold in the US in [----]. Today over [--] billion Oreos are sold. Fascinating. Oreo sales keep going up even though a lot of people have taken weight loss drugs. The Oreo substitution diet begins"
X Link 2025-08-11T23:34Z [---] followers, [---] engagements
"Great writeup Two thoughts 1) what if we dosed organic lithium well below the human-equivalent dose from mice studies to try to ward off Li-tox AEs would it do anything 2) serum Li is not equal to brain Li. What chemistry tricks can we do (packaging/targeting/composition) to shift more to brain"
X Link 2025-08-13T21:31Z [---] followers, [--] engagements
"After selling out not too long ago burry has gone bullish on united healthcare asml regeneron handful of chinese companies and a few other names. I did not expect burry to go long when market is roaring A lot of people are calling market top atm with ath in spy/qqq AI revenues being a product of vc paying indirectly for compute inflation etc BREAKING 🚨 Michael Burry just bough United Health too He just disclosed owning up to $100M of calls on $UNH https://t.co/vrfVbKpLq4 BREAKING 🚨 Michael Burry just bough United Health too He just disclosed owning up to $100M of calls on $UNH"
X Link 2025-08-15T02:09Z [---] followers, [---] engagements
"What happened with $PGEN has happened many times before for other orphan drugs. This is not a reflection of the current administration or Vinay Prasad. Seeing a lot of things about Prasad's quote shown below. He is right for rare severe diseases where no decent treatment currently exists. And this is not something that is novel its been done by previous directors / administrations too. Prasad has tweeted things like "RCT or STFU" but this stance is one that directors before him have taken and reflect the consensus opinion of drug evaluators in rare diseases"
X Link 2025-08-15T16:16Z [---] followers, [---] engagements
"Is $REPL going to get approved given $PGEN decision by Prasad"
X Link 2025-08-16T16:31Z [---] followers, [---] engagements
"In @A_May_MD pitch on $abvx he points out obefazimod inverse dose relationship. It should not have made people skeptical for [--] reasons: 1) rna dynamics make this plausible 2) in protein-receptor space this is true of inverse agonist drugs. Ofc hindsight is 20/20"
X Link 2025-08-17T01:07Z [---] followers, [----] engagements
"The survival curves of $ELTX mutant KRAS vaccine as adjuvant therapy in PDAC published in @NatureComms skim the median. The OS stays just above [--] from [--] to [--] months and a single event occurs dropped OS to zero (after [--] patients are censored). A similar story occurs with panel B where relapse free survival hovers around 60% from [--] to [--] months (2 patients get censored) one event moves it below 50% and the remaining patients get censored"
X Link 2025-08-18T17:33Z [---] followers, [---] engagements
"@varma_ashwin97 @DrSamuelBHume Clever hadnt heard of this before. Makes sense though. Could get pio approved if obesity-related endpoints are flat (mace obesity-related cancer). Obviously pio helps a1c"
X Link 2025-08-19T18:42Z [---] followers, [--] engagements
"People know glp1 = nausea etc. so if they dont have then they want to drop out to get drug/other glp1 perhaps some other way. How to drop out Maybe state AE and leave Were these marker/imaging/clinician witnessed (like you say you are having diarrhea but did I see it physically happen this is outlandish but captures the vibe) I have no position in viking. If we assume the above to be true then it raises the following point - its tough to evaluate any placebo arm ever and we can always read the placebo tea leaves however we want"
X Link 2025-08-19T20:48Z [---] followers, [--] engagements
"@AnilMakam Metformin and GLP1 curves are basically the same right So not magic just that metformin and glp1 both superior to insulin. Not mechanistic difference because insuling tx patients have more severe disease and other RF for cancer"
X Link 2025-08-22T16:10Z [---] followers, [---] engagements
"$IMRX up 65% since this tweet 4d ago Can dual MEKi and mutant KRASi add value over KRASi in PDAC (virtually all PDAC has a KRAS mutation) Atebimetinib (orphan drug status) is MEKi ($IMRX) and many KRASi approaches are approved (Sotorasib and Adagrasib) or in the works (Revolution Medicines). Would it make sense https://t.co/htKA8zUCkG Can dual MEKi and mutant KRASi add value over KRASi in PDAC (virtually all PDAC has a KRAS mutation) Atebimetinib (orphan drug status) is MEKi ($IMRX) and many KRASi approaches are approved (Sotorasib and Adagrasib) or in the works (Revolution Medicines). Would"
X Link 2025-08-25T16:03Z [---] followers, [---] engagements
"$PRAX is mispriced it should be worth at least 3B in market cap. Currently it sits at 1B market cap. Why Ulixacaltamide drug for essential tremor won't work but that's fine. Assign zero value to that asset. Relutrigine will work for SCN2A and SCN8A epilepsy but ALSO broad DEE too. Assume [------] total patients 20% max penetration slow linear ramp to peak penetration over [--] years $10k/pt/year discount rate at 5% annually. NPV of [--] years of sales is 3.6B. But that's not all. They have elsunersen which treats an ultra-rare disease of SCN2A GoF epilepsy patients. Conservatively throw a 200M NPV"
X Link 2025-08-27T21:21Z [---] followers, [----] engagements
"@KobeissiLetter Stock buyback at 4T valuation 🤯 Jensen feels they are undervalued at this point in the AI cycle"
X Link 2025-08-27T21:30Z [---] followers, [----] engagements
"@anssiapina 1) ok interesting then pts can recieve tx immediatet 2) agreed But does the extension of life come with any symptom improvement. I think either way its a choice for each pt to make based on their desires/wants/needs"
X Link 2025-08-28T13:51Z [---] followers, [---] engagements
"Two books I've been meaning to read for a while: 1) george soros: a life in full and 2) the alchemy of finance"
X Link 2025-08-28T15:23Z [---] followers, [---] engagements
"@ktkadakia @bostonsci @JACCJournals Woah incredible stuff Great to see boston scientific take into consideration research to make changes to their products"
X Link 2025-08-29T21:04Z [---] followers, [--] engagements
"@SinghJyotirmai What about the tariffs though How did US tariff announcement change India-China trade"
X Link 2025-08-29T21:23Z [---] followers, [--] engagements
"Sangamo ($SGMO) needs funding to continue. They have like [--] months left. They are coming to Wells Fargo and Cantor conferences in the next few days because they are either finalizing a deal or they want to drum up deal interest. I'm not sure which way its going to go. The thing tipping me to lean 55-45 finalizing a deal is that they have good Fabry data and they are still hiring actively. Dead companies don't hire. I could be wrong and this will serve as a public scorecard"
X Link 2025-09-03T02:12Z [---] followers, [---] engagements
"@princetongb What do you recommend I read to orient myself to macro stuff. I don't understand it well"
X Link 2025-09-03T18:29Z [---] followers, [--] engagements
"@RNAiAnalyst Im just confused why they keep opening job positions. Is it a headfake"
X Link 2025-09-04T19:21Z [---] followers, [--] engagements
"(5/) Areas for improvement Data data data High-quality multi-modal data curated for the problem at hand is the key. Architecture-wise how do we capture long distance informatio losing local detail Is context length the answer. Probably not. Local signal degrades as context length scales. And 1M nucleotides (largest length to date) is not even near chromosome level yet. Optimization environment: Can we set up a reinforcement learning environment for biology a la Prime Intellect Hub. Benchmarks: we need consensus datasets + metrics. Each DNA FM paper evaluates on different datasets and uses"
X Link 2025-09-04T22:28Z [---] followers, [--] engagements
"(6/) Applications of DNA Models: [--]. Designing state-specific promoters for CAR T AAV gene therapy and ddRNAi drugs [--]. Discovering new disease-causing targets through in silico mutagenesis [--]. Resolving variants of unknown significance [--]. Biosecurity These ideas are explored here: https://www.aditharun.com/p/dna-foundation-models https://www.aditharun.com/p/dna-foundation-models"
X Link 2025-09-04T22:28Z [---] followers, [---] engagements
"I think the ACMG guidelines are reasonable (Richards et al [----] Genetics in Medicine This evidence framework is what Ambry/Foundation/Myriad and others are judged against to afaik. They have a section on in silico prediction tools and write "These are only predictions however and their use in sequence variant interpretation should be implemented carefully. It is not recommended that these predictions be used as the sole source of evidence to make a clinical assertion". I agree DNA models or any other such model is a prediction at the end of the day. Table [--] attached is their classification of"
X Link 2025-09-06T18:59Z [---] followers, [---] engagements
"Im noticing weird behavior in options bids in low volume small market cap stocks. I think its algo trading / bot behavior. I place bid slightly above current bid I refresh and I am only person with said bid since it shows volume of [--] at that price [--] mins later I see [--] other bids at that higher price I cancel my bid now within 30s i see [--] bids spread across multiple premiums all slightly below my cancelled bid"
X Link 2025-09-12T16:29Z [---] followers, [---] engagements
"@GeneInvesting @Clarkitect80 From exhibit [----] filed w SEC in 2020"
X Link 2025-09-12T17:00Z [---] followers, [--] engagements
"@GeneInvesting @Clarkitect80 Plus minus a B or two"
X Link 2025-09-12T17:15Z [---] followers, [--] engagements
"It is incredible that GLP1s work for reducing cravings with evidence in alcohol use disorder opioid use disorder and cannabis use disorder. A little peptide that comes from the proglucagon gene primarily made by pancreatic and gut cells (and to a lesser extent in the brain) works like magic. It also fixes type [--] diabetes prevents heart attacks corrects excessive weight decreases incidence of many cancers and Alzheimer's disease. A GLP1-like medicine for preventing the habit circuit that leads to endless social media consumption would be incredible. Downstream consequences of accomplishing"
X Link 2025-09-12T21:27Z [---] followers, [----] engagements
"@3primeAnalytica @chris342222 What do you think about their value heading into rate cuts and possible inflation"
X Link 2025-09-15T18:14Z [---] followers, [--] engagements
"$qure however relies on AAV to treat Huntingtons Exclusive: Biogen is exiting AAV gene therapy joining a growing list of pharmas stepping back from the costly risk-laden delivery tech. https://t.co/lTVmPzciLS Exclusive: Biogen is exiting AAV gene therapy joining a growing list of pharmas stepping back from the costly risk-laden delivery tech. https://t.co/lTVmPzciLS"
X Link 2025-09-27T01:05Z [---] followers, 12K engagements
"ALS drug seemed to have a substantial OS benefit when comparing to external propensity-matched control arm (baseline characteristics pretty similar) but any benefit disappeared when comparing to placebo in intention-to-treat. This is important given the recent news about $QURE. They showed massive treatment benefit in Huntington's disease patients compared to an external control arm. This effect may disappear with placebo (or it may not and that's wonderful for patients) Note that $QURE's high-dose treatment arm that showed efficacy was n = [--] versus the ALS trial treatment arm that had n ="
X Link 2025-09-27T19:12Z [---] followers, [----] engagements
"OV329 does not accumulate in brain So if you give [--] mg/kg/d of the $ovid drug it never accumulates in the brain.so any on-target activity is transient inhibition Per poster abstract "Brain retina eyeball and blood samples were collected at [--] hours after initiation of SC infusion." Maybe I've missed something obvious here"
X Link 2025-09-29T22:45Z [---] followers, 14.8K engagements
"@JusticeforB_MD @ElicioTx Agreed Competitive landscape of 1L and 2L pancreatic cancer is poised to change soon $imrx meki in metastatic pancreatic cancer $rvmd pan-kras codon [--] inhibitor in metastatic pdac $eltx kras amphiphile vaccine in locoregional"
X Link 2025-09-30T02:59Z [---] followers, [---] engagements
"Has good news been announced for any public biotech company when they guided for a specific quarter missed that deadline completely and did not put out any release/commentary offering an explanation or postponement Re: $ovid q3 ph [--] topline"
X Link 2025-10-01T18:49Z [---] followers, [----] engagements
"Announced today: Evolocumab (PCSK9i) met both primary endpoints (time to first CV event time to CV event or ischemia-driven arterial revascularization) in VESALIUS-CV. Amgen positioning PCKS9i + lipid lowering therapy as the basis for primary prevention of CV disease. This trial can help inform future strategies for primary prevention of CV disease. Evo is given subq q2/q4 weeks so opportunity exists for a therapy that needs to be given less often. ANGPTL3 inhibition and remnant cholesterol targeting drugs could play a role. Atherosclerotic CV disease often co-occurs with other chronic"
X Link 2025-10-03T02:44Z [---] followers, [----] engagements
"@TCellEngager @jesse_brodkin If EMA rejects then I'm not sure what the company has. Schizophrenia pipeline doesn't make sense Fragile X MoA is totally distinct from Sigma-1R so that won't work etc. But then again stock went up 10% today on schizophrenia data announcement. so im confused"
X Link 2025-10-04T00:26Z [---] followers, [---] engagements
"I need to think more about the pdac clinical landscape. I like their approach and data has been clean in that no tricks ive discovered on reporting information. Small n which is the main concern but that will be solved w ph3. The extent to which low dose chemo might be driving the signal is another concern before I reach a conclusion. Fwiw rn i do not think low dose can have this sort of effect"
X Link 2025-10-06T00:56Z [---] followers, [--] engagements
"@hannibalspeaks sports betting but better"
X Link 2025-10-12T21:54Z [---] followers, [--] engagements
"What's a good book to read on Lee Kuan Yew"
X Link 2025-10-13T19:04Z [---] followers, [---] engagements
"Nicotine promotes a chronic inflammatory state promote pro-inflammatory cytokine release and suppresses immune responses against Gram-negative bacteria by impairing mitrochondrial function. This acts as a nidus for periodontal disease gum recession and tooth loss caused primarily by gram-negative anaerobic bacteria. Nicotine-free pouches (like Ultra) try to deliver the same hit without the consequences of nicotine. There is no free lunch but maybe its a good discount lunch. I have no idea I've never tried them. But here's what's in them: [--]. Paraxanthine Paraxanthine (17-dimethylxanthine) is"
X Link 2025-10-14T12:18Z [---] followers, [---] engagements
"I had a toilet issue today. Normally Id call someone but this time I put in my AirPods opened ChatGPT voice mode and walked through it step by step. I reasoned through what I was seeing what the parts seemed to do and what I thought should happen with the AI. I had zero experiencejust knew there was a chain and a fill valve. But with voice mode and occasionally flipping to video I got 90% of the way there. The last 10% I figured out myself. The big thing is that ChatGPT voice allowed me to do something that I wouldve never felt comfortable doing before"
X Link 2025-10-14T23:51Z [---] followers, [---] engagements
"Im really intrigued about holding $jepi or similar high yield equities. The primary risk is they cut dividends when cash flow drops"
X Link 2025-10-16T19:32Z [---] followers, [---] engagements
"The states w highest median income tend to have higher credit card debt Based on eyeballing the data fyi Americans are drowning in credit card debt: US household credit card debt hit a record $1.33 trillion in August. The combined average credit card balance across all age groups is now at whopping $10668 PER household. States with the largest average debts in Q2 [----] were https://t.co/QSZgTBa8XB Americans are drowning in credit card debt: US household credit card debt hit a record $1.33 trillion in August. The combined average credit card balance across all age groups is now at whopping"
X Link 2025-10-16T23:31Z [---] followers, [---] engagements
"@RyanSTTrader I dont understand this. Spy is flat on the week and sentiment is extreme fear. I guess forward looking anticipation of drop is the consensus"
X Link 2025-10-17T15:30Z [---] followers, [--] engagements
"Pregnant women with HIV on treatment and with CD4 [---] may have significantly lower odds of having a preterm birth if given tmp-smx during pregnancy. 2% of births in the trimethoprimsulfamethoxazole group were preterm as compared with 14% in the placebo group. The table below shows the results for HIV positive women in this study - better in many categories when given tmp-smx (cotrimoxazole). These are exploratory analyses not powered to make claims. We need an RCT specifically looking at this population. This study finds that pregnant women given TMP-SMX 80mg-400mg BID after [--] weeks"
X Link 2025-06-09T15:56Z [----] followers, [---] engagements
"Google has an incredible amount of data to leverage for LLM development. They fine-tuned a LLM for CRISPR-GPT based on [--] years of a public Google Discussion Group on CRISPR gene-editing. Now of course they can't use X which is probably the best real-time source of information. Interestingly real-world research questions performance drops 20% when adding in this source of data. Perhaps the fine tuned model over-indexes on the forum which contains a lot of wrong deep technical CRISPR questions (because scientists post what they cannot figure out / confused about)"
X Link 2025-08-06T00:07Z [---] followers, [---] engagements
"Tbh I thought $vyne [---] results would be good. I was surprised when they were so bad. But I think its something to dissect and learn from. On my todo list. A post-mortem on this could be very informative"
X Link 2025-08-06T12:00Z [---] followers, [---] engagements
"Zero shot Evo2 (AUROC 0.65ish) vs. ProteomeLM-Ess fine-tuned for essentiality (AUROC 0.9) for predicting gene essentiality. However ProteomeLM-Ess is only for protein coding genes. What is Evo2 zero-shot performance subset to only protein coding genes With fine-tuning Evo2 can be as good as ProteomeLM-Ess"
X Link 2025-08-06T19:10Z [----] followers, [---] engagements
"Can dual MEKi and mutant KRASi add value over KRASi in PDAC (virtually all PDAC has a KRAS mutation) Atebimetinib (orphan drug status) is MEKi ($IMRX) and many KRASi approaches are approved (Sotorasib and Adagrasib) or in the works (Revolution Medicines). Would it make sense to combine the two. We can look to Sotorasib (KRAS G12Ci) + Trametinib (MEKi) in heavily pre-treated NSCLC patients (median [--] lines of prior tx) had ORR of 14% (prior KRAS G12Ci given n = 7) and 9% (KRAS G12Ci naive n = 11). In CRC patients ORR of [--] in prior KRAS G12Ci (n = 3) and 20% in KRAS G12Ci naive (n = 15)."
X Link 2025-08-20T17:30Z [----] followers, [----] engagements
"Yeah. All patients had no cancer at start and some developed cancer. If we think about CRC the HR [---] 0.46-0.64 doesn't really make sense. They propensity matched on a number of criteria including adverse socioeconmic status and prior colonoscopy. We know T2DM patients requiring insulin have more severe disease than those who do not require insulin. We have no idea what the A1C levels are between groups in this study. We also know routine colonoscopies can detect and treat polyps that lead to CRC. We do not know if patients with T2DM on insulin had different colonoscopy rates following"
X Link 2025-08-22T16:52Z [----] followers, [--] engagements
"Sangamo is an interesting company. Summary of their work and my thoughts on their specific clinical programs. 1) They have a brain specific AAV that delivers zinc fingers (ZF) partnered with Lilly/Genentech/Astellas. It has not been tried in humans yet only primates. But in primates it seems to work. If partner co's can't find something good no further milestone payments. I expect this to be the case. 2) They have primate data on IV ZF + brain AAV in prion disease. I'm skeptical of prion results (see reply). And $SGMO has an intrathecal ZF + brain AAV product for idiopathic small fiber"
X Link 2025-08-28T04:27Z [---] followers, [----] engagements
"$SGMO ASGCT talk presented prion primate data. In primates (n=2) prion transcript 25% of full disease in different brain regions. Does 75% reduction in NEW prion transcript matter Patients with prion disease progress quickly receive testing for all sorts of other things then get ruled into this diagnosis. So by the time patient arrives ready for treatment they will have lots of prion protein already misfolded and aggregated. This drug slows rate of NEW misfolded proteins joining the aggregated by 75%. Will that translate into regression of patient symptoms or even stabilization Probably not"
X Link 2025-08-28T04:27Z [----] followers, [----] engagements
"@yusufhameed @MofeekS Appreciate the viewpoint from pain perspective. Did not know"
X Link 2025-08-28T14:58Z [----] followers, [--] engagements
"This is really cool Interesting that all post-tx muts were present pretx suggesting the tx acts exerts selection pressure but is not a wbc mutagen. Do you think it becomes a mutagen if chemo given for many more cycles If Im interpreting the vaf change per month plot on each tx at [--] year a 12% median increase in vaf is expected on chemo What % increase is considrered pos selection in your km curves"
X Link 2025-09-03T13:11Z [----] followers, [--] engagements
"Focal onset epilepsy has a few key players at the moment ( $RAPP $XENE $PRAX) but it is my belief that $PRAX is undervalued and has a lot of upside. Described my thoughts here: If $PRAX executes on rare disease epilepsy alone it will justify a 3B valuation imo (3x upside from here) and I feel that is almost a sure thing. https://www.aditharun.com/p/praxis-epilepsy-drugs-will-work-but https://www.aditharun.com/p/praxis-epilepsy-drugs-will-work-but https://www.aditharun.com/p/praxis-epilepsy-drugs-will-work-but https://www.aditharun.com/p/praxis-epilepsy-drugs-will-work-but"
X Link 2025-09-20T17:55Z [----] followers, [---] engagements
"@adamcifu illustrating Goodhart's Law in medical school admissions "Those uninterested in research sacrifice time during which they could be doing what they love to fill third-tier journals with drivel" Goodharts Law and Medical School Admissions Sensible Medicine Oct [--] 2025"
X Link 2025-10-11T13:15Z [---] followers, [----] engagements
"$PRAX up 224% today The valuation pre-jump today supported a 2-3x upside just off epilepsy portfolio and it came with a flyer on the essential tremor drug And today they announced Positive Topline Results from Two Pivotal Phase [--] Studies of Ulixacaltamide HCl in the Essential3 Program for Essential Tremor I detailed this in the quoted post Have not studied the data or results will look into it. My initial thoughts are that Im quite surprised the ET trial was a success $PRAX is mispriced it should be worth at least 3B in market cap. Currently it sits at 1B market cap. Why Ulixacaltamide drug"
X Link 2025-10-16T14:18Z [---] followers, [----] engagements
"Regeneron 👀 on the heels of NEJM DB-OTO results The [--] selected: Makary says more vouchers set to be announced in the coming months as well. https://t.co/bOxlnTuiE9 The [--] selected: Makary says more vouchers set to be announced in the coming months as well. https://t.co/bOxlnTuiE9"
X Link 2025-10-16T22:11Z [----] followers, [---] engagements
"Lp(a) drugs are coming soon Pelacarsen (Ionis) Lepodisiran (Eli Lilly) Olpasiran (Amgen) Zerlasiran (Silence Therapeutics) All with positive Phase [--] and in phase 3; All subq injections Muvalaplin (Eli Lilly) Positive Phase [--] announced. Oral route of administration. It's happening Lipoprotein(a) testing is going to skyrocket once it's possible to treat it. https://t.co/aFhMiA4Zu6 It's happening Lipoprotein(a) testing is going to skyrocket once it's possible to treat it. https://t.co/aFhMiA4Zu6"
X Link 2025-10-17T14:15Z [---] followers, [----] engagements
"1/3 Men's testosterone levels did not plummet over the past decade like many wellness podcasts would have you believe. we just switched the measurement assay from immunoassay to mass spectrometry When we switched assays we observed a 100% increase in self-reported healthy men with low testosterone. The threshold needed to be updated when the assay changed. Arun AS Durant TJS El-Khoury JM Krumholz HM. Re-evaluating the threshold for low total testosterone. Clinical Chemistry 2025"
X Link 2025-10-20T19:36Z [---] followers, 406.7K engagements
"3/3 I found this post where @eryney_ok and @mu_anand rightly pointed out that changes in measurement tools have led people to falsely believe their T is low"
X Link 2025-10-20T19:36Z [---] followers, 13K engagements
"Druckenmiller offering a masterclass in just over an hour. Will need to re-listen to this again. Takeaways: - Discipline. No matter how good you are once your ego inflates its over - There are many ways to make money in the market the biggest thing is to have a system - Think deeply about where and how you went wrong. People who dont spend a lot of time talking about their failures and focus on their wins are negative signal. - Historical correlations and patterns break when black swan events happen. So standard risk models break during these events and that is exactly when you want risk"
X Link 2025-10-20T23:15Z [---] followers, 305.2K engagements
"@princetongb $argt"
X Link 2025-10-21T15:44Z [----] followers, [---] engagements
"@TheIngated I am very interested in figuring out the health effects of microplastics. First places I might look for an association are gut issues and crc in young adults. The hard part is finding microplastic levels in humans indexed with outcomes"
X Link 2025-10-22T02:13Z [---] followers, [----] engagements
"The entire paper is only [--] page and its attached as an image to the second tweet of my original post. But yes we moved from immunoassay to mass spec and kept using the immunoassay threshold the whole time. The two methods perform largely the same but are not exactly the same. Im not sure that other trace molecules need to have their threshold re-calibrated. If the same story was found for another molecule then yes. But I see it as case by case"
X Link 2025-10-22T02:26Z [---] followers, [--] engagements
"be a fund manager send a guy to catalogue jackets salespeople at each b2b saas company wore at conferences sales people splurge on jackets when deal flow is up its a discretionary purchase average jacket price for company is a leading indicator of earnings Cheeky work love to see it"
X Link 2025-10-24T02:51Z [---] followers, [---] engagements
"Interesting hypothesis for anktiva (il15 agonist) in cancer treatment will need to look into data. Immediate thought is that we know il2 boosts nk and t cell counts but it doesnt work well / has many adverse effects Working to explain to the world how and why Bioshield works. T cells and NK cells kill cancer thats irrefutable. Low NK cells and T cells means cancer grows thats irrefutable. So if there is a drug available to boost NK and T cells then there is better chance to kill https://t.co/9nbHmD2coL Working to explain to the world how and why Bioshield works. T cells and NK cells kill"
X Link 2025-10-24T19:01Z [---] followers, [---] engagements
"@AnilMakam @NEJM @tabuckley_ @arjunmanrai @Gurpreet2015 Agreed What do you think the role of the largely non-procedural physician is in the age of AI My thoughts: Feedback / thoughts appreciated https://open.substack.com/pub/aditharun/p/what-does-a-doctor-do-in-the-ager=4do42&utm_medium=ios https://open.substack.com/pub/aditharun/p/what-does-a-doctor-do-in-the-ager=4do42&utm_medium=ios"
X Link 2025-10-24T20:22Z [---] followers, [---] engagements
"Will Q4 $TECX Ph1b topline readout of long-acting relaxin (TX45) in pulmonary hypertension (PH) associated with HFrEF be positive Alise (CEO $TECX) said in her Morgan Stanley talk that data is to be expected 2H of October. First by "positive" in this Ph1b I mean two things: 1) its safe in this patient population and 2) it moves the hemodynamic parameters in the right direction. Down the line positive will also mean reduction in symptoms and increased [--] minute walk test but this study measures hemodynamic parameters up to [--] hours after single dose of TX45. So the parameters to judge TX45 in"
X Link 2025-10-24T21:05Z [---] followers, 12.8K engagements
"@plainyogurt21 ok good to know I heard alise mention it but couldn't find it on azn materials in my search probably just missed something"
X Link 2025-10-24T22:04Z [---] followers, [---] engagements
"@plainyogurt21 can you clarify what you mean by "if you believe the pathology is actually different where the heart can handle additional CO and the pathology is truly RHD and PH""
X Link 2025-10-24T22:07Z [---] followers, [---] engagements
"Ok i think I understand. Lilly looked at acute decompensated hf and tecx/azn is stable hf like you pointed out. And that is the difference imo. Stable means pts have physiologic reserve and thats what counts. Once youre decompensated all bets are off. To me its kind of like saying why doesnt my chemo work in metastatic disease well maybe you should have given it in earlier disease Now your second point on pvr is one i dont agree with. Tecx showz pvr [--] and pvr =2. Pvr declines more when baseline pvr high but that is expected in cpcph since right sided pressures are affected. In ipc pvr is"
X Link 2025-10-25T01:51Z [---] followers, [---] engagements
"My favorite quote from The Technologic Republic is highly relevant to engineers scientists and physicians: be multidiscplinary. Study things from different fields and apply them. That's what da Vinci Gutenberg and Brunelleschi and others did because they had to"
X Link 2025-10-26T22:08Z [---] followers, [---] engagements
"AZN ran study in patients with any EF so all comers but if my hypothesis is that it works regardless of EF then we will see an effect. The other relaxin trial papers have supplemental analyses by EF and there was no difference in effect sizes. This is another point that makes me feel lile what happens in hfpef will happen in hfref. Now given the different trial pop in azn and tecx compared to lilly this statement requires a leap of faith because of the different trial pops. I think its a reasonable leap. Yes i do think a chronic stable class II or III HF patient has orders of magnitude more"
X Link 2025-10-27T13:29Z [---] followers, [---] engagements
"Plain pulmonary vasolidation is no good in hfpef as weve talked about. The relaxin case is gentler vasodilation lusitropy help and with long term use anti-fibrosis - as weve both mentioned. Hard to say if that comes to fruition. I can say that the update in q4 is only hemodynamics not functional improvement so thats all we can care about for this upcoming one Safety is key agreed. Well hopefully be able to get data in the next [--] days"
X Link 2025-10-27T14:09Z [---] followers, [---] engagements
"Who makes $NTLA's LNP Is it Novartis I believe MoA of LNP is via ApoE on hepatocytes for ATTR programs"
X Link 2025-10-27T15:06Z [---] followers, [---] engagements
"Recap of $NTLA call Oct [--] 2025: Male patient with ATTR-CM given dose of $NTLA gene therapy on Sep [--] [----]. AST and ALT [--] times upper limit of normal and bilirubin [--] times upper limit of normal. Voluntary pause. Not mandated by FDA. Patient complained of abdominal pain and was evaluated in ED where labs were drawn. AST ALT and bilirubin were elevated. That happened Oct [--] [----]. (*) Clinically the patient is stable per management playing cards with nurse as of last night. Management says "would meet the traditional definition of Hy's Law". Hy's law is a threshold at which fatal drug-induced"
X Link 2025-10-27T15:33Z [---] followers, [----] engagements
"@JusticeforB_MD SCD Pain crises auto splenectomies and impaired immunity VOCs increasing risk for many things is not solved by current treatment"
X Link 2025-10-27T19:50Z [---] followers, [--] engagements
"AZN sets the expectations for relaxin in PH thats the bigger trial. If AZN is positive then HFrEF TECX will be positive. If for some reason HFrEF TECX gets announced first then I expect AZN primary endpoint of PVR to hit. Safety in TECX is single dose but AZN is repeated dosing so safety from AZN is a more accurate measure of safety of relaxin. Your theory re: cpcPH is plausible and just to add showed Cpc is the dominant type of PH in HFrEF and HFpEF with like 80% of cases being that way in either etiology of HF. Great discussion really enjoying it https://pubmed.ncbi.nlm.nih.gov/26181215/"
X Link 2025-10-27T23:07Z [---] followers, [---] engagements
"@mtheory11bio Optimistic they'll release it any day now"
X Link 2025-10-28T00:48Z [---] followers, [---] engagements
"Lots of talk about Hy's law today after Intellia's Grade [--] adverse event disclosure. Some context into the test characteristics of Hy's law: Hy's Law (ALT or AST 3x ULN & TBili 2x ULN & exclusion of other causes) predicts a 10% risk of death or liver transplantation among affected patients. In a large US managed care cohort Hy's law had a specificity of 92% and a PPV of 2% for acute liver failure among drug-induced liver injury cases. Most patients who get drug-induced liver injury do not progress to acute liver failure. Without the patient's chart its hard to say which way the scales might"
X Link 2025-10-28T00:57Z [---] followers, [---] engagements
"I do not understand the psychology of the linkedin warrior"
X Link 2025-10-28T02:47Z [---] followers, [---] engagements
"@3primeAnalytica regn with strong q3 earnings Nice call"
X Link 2025-10-28T12:32Z [---] followers, [--] engagements
"$VYNE BET inhibitor for nonsegmental vitiligo also canned earlier this year (negative trial results) Another project gone: $INCY's BET inhibitor INCB57643 https://t.co/1HDrjrBmv8 Another project gone: $INCY's BET inhibitor INCB57643 https://t.co/1HDrjrBmv8"
X Link 2025-10-28T12:34Z [---] followers, [----] engagements
"gene therapies have been derailed by grade 3/4 AEs before and this is a risk to putting money in gene tx companies. Ntla had one asymptomatic patient before which made me think huh thats interesting but the second one soured my appetite. I certainly didnt predict the event and am curious what the official explanation is and how it may have happened biologically"
X Link 2025-10-29T11:20Z [---] followers, [--] engagements
"Patiently waiting for AZN or TECX updates Alise CEO of TECX said 2H of October which means by Friday EOD Will Q4 $TECX Ph1b topline readout of long-acting relaxin (TX45) in pulmonary hypertension (PH) associated with HFrEF be positive Alise (CEO $TECX) said in her Morgan Stanley talk that data is to be expected 2H of October. First by "positive" in this Ph1b I mean two things: 1) https://t.co/kHLa0Y0Xk4 Will Q4 $TECX Ph1b topline readout of long-acting relaxin (TX45) in pulmonary hypertension (PH) associated with HFrEF be positive Alise (CEO $TECX) said in her Morgan Stanley talk that data is"
X Link 2025-10-29T14:30Z [---] followers, [---] engagements
"Recently med school classmate (ex-Amazon SWE) and I wrote up some slides on how to think about using AI in medicine with example cases to work through. Prompt engineering evaluating outputs workflow in clinical reasoning etc. Taught it as an interactive session to M2s and will do it for M1s in the coming weeks. It is great to hear and see how they are thinking about and use it - I learn a lot. How we use AI in medicine will continue to evolve rapidly so teaching a specific ways of use may only be valuable for a short time. Instead principles of AI how models are trained and bayesian reasoning"
X Link 2025-10-29T15:45Z [---] followers, [---] engagements
"The TECX APEX Ph2 trial (topline 2026) and AZN Ph2 trial (expected readout any day now) have the exact same setup except AZN is recruiting PH-HF patients regardless of LVEF and independent of ipcPH/cpcPH designation. TECX seems to be doing a good job working to find the right group of patients that will respond. But by in large if AZN Ph2 succeeds then TECX Ph2 should succeed too. TECX APEX Ph2 trial (topline expected 2026) is evaluating chronic treatment in [---] PH-HFpEF patients enriched for PVR at or above [--] (70% of population) [--] min walk test is secondary endpoint primary endpoints are"
X Link 2025-10-30T15:08Z [---] followers, [---] engagements
"$REGN Q3 earnings call summary - 3.8B total revenues (up 1% YoY) diluted net income per share of [-----] 1.3B net income. 198M of profit share royalty and licensing. 1.3B R+D expenses (mid-teens increase in expenses in [----] to be expected given trials upcoming). Cash less debt of 16B dollars. - First [--] months of [----] repurchased 2.8B of shares (most for any fiscal year for REGN) anticipate returning 4B in shareholders in dividends and repurchases in [----]. - Dupixent WW net sales 4.9B. 29% YoY growth. 3.6B sales in US 28% YoY growth. Leads market in all established indications. Pulmonogists"
X Link 2025-10-30T16:19Z [----] followers, [---] engagements
"@3primeAnalytica had a great summary. It is much more succinct than mine and pulled in key quotes from the call would recommend https://x.com/3primeAnalytica/status/1983273052843475443 $REGN what a quarter 🟢12% on earnings & updates: 📍Rev at $3.75B (vs. $3.59B expected +1% YoY). ✅ 📍Non-GAAP EPS of $11.83 exceeded forecasts ($9.64 est.) despite IPR&D charges. GAAP net income rose 9% to $1.46B. ✅ Key drugs: 📍Dupi (what a monster) at $4.86B (+26% https://x.com/3primeAnalytica/status/1983273052843475443 $REGN what a quarter 🟢12% on earnings & updates: 📍Rev at $3.75B (vs. $3.59B expected +1%"
X Link 2025-10-30T16:20Z [---] followers, [---] engagements
"@drug_smolecules Anticipating AZD3427 announcement Nov [--] Also where did you find the AZN doc you screenshotted in the original post"
X Link 2025-10-30T19:09Z [---] followers, [---] engagements
"Thanks for sharing Shelby My guess would be that non-platform companies are more likely to succeed because they are focused on only one thing. This finding is quite interesting. How many platform companies actually brought [--] asset to successful phase [--] A lot of public biotech companies bill themselves as platform companies to investors and have a platform "up on their site" but there is maybe [--] or [--] total assets"
X Link 2025-10-31T13:46Z [---] followers, [--] engagements
"@financebully @HOThomasWPhelps Pbo adjust each metric is my view"
X Link 2025-10-31T15:00Z [---] followers, [--] engagements
"@HOThomasWPhelps Ok wow nice"
X Link 2025-10-31T15:35Z [---] followers, [---] engagements
"If curing inherited deafness is not incredible then I don't know what is. A short thread on a beautiful paper 1/3: Otoferlin (synapse transmission of hair cells in ear) deficiency results in congenital deafness. A single intracochlear infusion in one or both ears of a AAV1 gene therapy made of OTO cDNA and hair cell-specific promoter allows these individuals to hear. - [--] of [--] patients can hear [--] dB - [--] of [--] patients treated in both ears achieved [--] decibels. Median response was [--] dB at [--] weeks. - [--] of [--] participants can hear soft speech with gene therapy alone ( [--] decibels) and 3"
X Link 2025-11-04T05:36Z [---] followers, [----] engagements
"What about the [--] non-responders ( [--] dB at [--] weeks) [--] eventually hit [--] dB at [--] weeks [--] was [--] years old at infusion. Patients [--] [--] and [--] had week [--] levels below [--] dB. However pt [--] eventually did recover. Pt [--] never responded. Pt [--] was [--] years old. Participant [--] (4yo M DB-OTO and implant concomitantly placed) had hearing improvements after [--] weeks but not to primary endpoint level of [--] dB. At week [--] with CI deactivated 100% accuracy at syllable patterns and 50% accuracy for 2-syllable words with no visual cues. Patient [--] the other [--] yo F had a barely above threshold response at week"
X Link 2025-11-04T05:36Z [---] followers, [---] engagements
"DB-OTO is safe. However note that there is roughly double the AE rate in Part A (one ear) compared to Part B (both ears treated). My guess is that the surgeons and surgical technique get better with more reps. Another hypothesis is that the cochlear implant actually comes with side effects given that [--] of the part A patients concomitantly received implants with gene therapy. [--] of [--] AEs were surgery related. [--] serious AEs: Grade [--] mastoiditis associated with CI in an ear not treated with DB-OTO and Grade [--] walking instability vestibular impairment (recent varicella vaccine). Both resolved"
X Link 2025-11-04T05:36Z [---] followers, [---] engagements
"Inflation is not 3% Tracking my own basket of goods YoY its more like 10%. I do not think wages are going up 10%. Imagine having dependents.only so far you can stretch a dollar. I remember seeing something about how hamburger helper sales have recently been trending up. not a good sign. - Fed rates are going down - Fed ending QT in Dec liquidity to go up - [--] y T-bill minus [--] y T-bill maturity flat at 0.4% last [--] months constant upward movement since [----]. Growth expected to be slow a bit but not drop too much - US Q2 GDP YoY 2% increase - GOOG R+D 36B https://t.co/XNv1pQfhmt - Fed rates are"
X Link 2025-11-04T18:00Z [---] followers, [----] engagements
"$TECX down 34% to $16.4/share since Ph1 PH-HFrEF readout. Likely sell the news and then concern over $AZN Ph2 PH-HF readout. Druckenmiller said at some talk I listened to that if the asset is trading down on no news then its a bad thing. I could argue here that there is news its just worry over AZN Ph2 and their perceived lateness in announcing results. AZN Q3 earnings 11/6 8a before market opens. The consensus on X is that AZN will announce Ph2 PH study results and/or offer guidance about the program. I'm sure an analysts will ask if not and management's words will be informative (or at"
X Link 2025-11-04T22:06Z [----] followers, [----] engagements
"$TENX is developing an oral formulation of levosimendan an approved drug in [--] countries (not the US though) for acute decompensated HF (improves symptoms reduce hospitalizations) for PH associated with HFpEF. IV levoseimendan has been tried in PH-HFpEF by TENX actually and the results were reported by Burkhoff et al JACC HF [----]. Notably it improved [--] minute walk test. It didn't move hemodynamics. PO levoseimendan in severe HF did not seem to meaningfully move QoL scores. Trial done by Orion maker of IV levoseimendan. However compared to placebo a net improvement of [--] points in MLHF QoL at"
X Link 2025-11-05T19:33Z [----] followers, [----] engagements
"Nice large cup of coffee and tuning into astrazeneca's earnings call. good start to the day"
X Link 2025-11-06T12:58Z [---] followers, [---] engagements
"Doordash a cheeseburger from McDonalds through chase credit card 23% APR Doordash is an IQ test. Do you want a cheeseburger for $27 or will you go pick it up for $6 Doordash is an IQ test. Do you want a cheeseburger for $27 or will you go pick it up for $6"
X Link 2025-11-06T16:37Z [---] followers, [---] engagements
"For context pluvicto made by Novartis is a radiotherapy that prolongs survival in metastatic prostate cancer. It was first approved in [----] and the label was expanded in [----]. Pluvicto is made of a prostate-specific membrane antigen (PSMA) binder linked to lutetium-177. Once given IV pluvicto binds cells that express PSMA and the lutetium-177 emits beta-minus radiation which induces DNA damage and cell death. This is targeted radioactive destruction of cancer cells by a small molecule similar to radioactive iodine used in treating thyroid cancer. In [----] pluvicto was approved for treating"
X Link 2025-11-07T01:46Z [---] followers, [----] engagements
"@buccocapital Do chipotle"
X Link 2025-11-07T04:07Z [---] followers, [---] engagements
"stochastic CRISPR-cas9 indels / frameshifts can lead to either toxic mRNA or an mRNA that is able to break through nonsense mediated decay and create either a toxic or immunogenic protein -- this is because of the specific region of the genome that ntla targets because of the specific crispr machinery ntla uses or just a class-effect in that any frameshift introducer runs this risk or perhaps some combination of the three. clarifying for myself"
X Link 2025-11-07T04:37Z [----] followers, [---] engagements
"Hacked together an app over the weekend (w a friend from hs) to track hair density from a selfie and applied to YC in [----] and got an email saying top 10% or maybe 5% can't remember It was born from an actual need: my hair was thinning. Basically I got frustrated with being told to try different things (XYZ shampoo oral minoxidil ABC supplement) but there being no actual good way to measure my hair density the metric I actually cared about. If you have an intervention you should have a measurement tool for it beyond "look in the mirror and go off vibes" or "take a photo and manually inspect"
X Link 2025-11-07T05:06Z [----] followers, 52.5K engagements
"@HairDAO_ Not building custom hardware. Can accept video output from whatever source built initially for standard iphone camera input"
X Link 2025-11-07T16:39Z [---] followers, [----] engagements
"I got that dog in me The dog in question:"
X Link 2025-11-07T18:51Z [---] followers, [---] engagements
"@houndcl This was my main concern pre data readout. But bezuclast is less toxic. Wonderful. Hard to predict. What biochemical clues did you see pre-today's data Trying to learn to look at biochem data more"
X Link 2025-11-10T16:32Z [---] followers, [---] engagements
"@doctorveera Great post What do you make of the bili spike a few days post tx"
X Link 2025-11-10T17:13Z [---] followers, [---] engagements
"Is there a good long thesis on tractor supply co out there $tsco politicians are buying and i like the company - sales staff in store is great and company is basically a proxy for agriculture industry in US Very reasonable PE. Not knowledgable beyond this"
X Link 2025-11-10T17:28Z [----] followers, [---] engagements
"@Ridingwaves69 Nobody is paying me can't say if the same is true for others I have no information there. You are long AVXL Please share thesis"
X Link 2025-11-10T19:23Z [---] followers, [---] engagements
"@3primeAnalytica @Regeneron what do you think about city therapeutics they have ASO for Factor XI advancing to clinic if I remember clinic with timeline of 2029"
X Link 2025-11-10T19:26Z [----] followers, [--] engagements
"agreed that there is an unmet need agreed that does not cause ARIA as much as current EMA-approved drugs do not understand your point on insolvency we don't treat most patients with AD with anti-amyloid therapy at baseline. it is shared decision making and certainly not stadnard of care. i also disagree i think most docs treating AD patients may not use once they read trial but I don't hold this point strongly given that it could change placebo is also safe . avxl doesn't beat the current donanemab results.and no mechanistic plausability unlike donanemab imo.also lots of weird stats /"
X Link 2025-11-10T19:55Z [---] followers, [--] engagements
"@adamfeuerstein If creating a bell labs was as easy as getting some cash together and hanging a sign that says bell labs we would have solved quantum computing by now Capital with no strings attached is a great start but like arena there are strings attached especially when softbank comes in"
X Link 2025-11-11T00:41Z [---] followers, [---] engagements
"I don't think you can create a Bell Labs just because you raised some money from sama and softbank. At some point (sooner rather than later) softbank is going to expect return. And research especially at the cutting edge is incredibly uncertain - discoveries are not churned out every quarter. Bells Labs worked because there was a company with a monopoly over an industry that funded its work without strings attached. After bell was broken up funding dried. softbank can only commit so much capital even if it is a couple billion or some crazy number. It's preferable that the lab be attached to a"
X Link 2025-11-11T01:14Z [---] followers, [----] engagements
"@Drlipid @Amgen Woah did not put [--] and [--] together. Thats pretty cool naming from them"
X Link 2025-11-11T01:30Z [----] followers, [---] engagements
"@Sanjusinha7 Just my vibes in terms of rank prioritizing targets. Is this not consensus in preclin drug dev To argue against myself chemo is broadly toxic and they are standard in many fields"
X Link 2025-11-11T17:19Z [---] followers, [--] engagements
"If true $nvo evoke will read positive We will know Dec [--]. Im not sure why but novo management said they didnt know or didnt want to know cant remember (someone please correct) Apparently the connection between GLP-1 and AD is not so far fetched. Cleveland Clinic did a real-world observation study and found that patients who initiated GLP-1s had 31% lower risk of developing AD. SGLT2i had 33% reduction. DPP4i was the control. https://t.co/xWOyJTWNkk https://t.co/FBMHGpAipI Apparently the connection between GLP-1 and AD is not so far fetched. Cleveland Clinic did a real-world observation study"
X Link 2025-11-12T03:20Z [---] followers, [---] engagements
"@avidresearch If reads positive big market unlock and maybe first real tx for AD"
X Link 2025-11-12T03:55Z [---] followers, [--] engagements
"@avidresearch Not for nothing but iirc burry disclosed 2% of portfolio (notional value) in pfizer calls Pfizer is a winner bc of metsera deal if novo reads out positive on evoke @avidresearch brought this up earlier today"
X Link 2025-11-12T03:57Z [---] followers, [--] engagements
"If true $nvo evoke will read positive We will know Dec [--]. Im not sure why but novo management said they didnt know or didnt want to know cant remember (someone please correct) Apparently the connection between GLP-1 and AD is not so far fetched. Cleveland Clinic did a real-world observation study and found that patients who initiated GLP-1s had 31% lower risk of developing AD. SGLT2i had 33% reduction. DPP4i was the control. https://t.co/xWOyJTWNkk https://t.co/FBMHGpAipI Apparently the connection between GLP-1 and AD is not so far fetched. Cleveland Clinic did a real-world observation study"
X Link 2025-11-12T03:58Z [---] followers, [---] engagements
"@therealRYC @NatureMedicine Theyre basically the same imo"
X Link 2025-11-13T03:51Z [----] followers, [---] engagements
"@therealRYC @NatureMedicine Interesting that TZP has accelerating market share and semaglutide is decelerating TZP has rate of weight loss but you basically end up in the same place wirh a long enough time horizon which is what matters for non-aesthetic reasons"
X Link 2025-11-13T03:53Z [----] followers, [---] engagements
"$MRSN up 210% $DAWN to acquire for 25/share with cash up to 30.25/share 🤯 tbh i didnt have this on my radar There is no good therapy for patients with metastatic TNBC who have many lines of prior therapy including at least [--] prior topo-1 ADC (like sacituzumab govitecan) median PFS is 1-3 months and median OS is 6-8 months. If patients are HER2 low (so not truly TNBC) they may get https://t.co/N0otXrfHby There is no good therapy for patients with metastatic TNBC who have many lines of prior therapy including at least [--] prior topo-1 ADC (like sacituzumab govitecan) median PFS is 1-3 months and"
X Link 2025-11-13T14:30Z [----] followers, [----] engagements
"@varma_ashwin97 @therealRYC @NatureMedicine @therealRYC and I reporting for duty @novonordisk"
X Link 2025-11-13T15:45Z [----] followers, [--] engagements
"I am surprised (and confused) how $avxl is still at a 320M market cap this morning"
X Link 2025-11-14T14:45Z [----] followers, [----] engagements
"@Banana_Oncology @jimcramer Inverse cramer is going to be wonderful for xbi"
X Link 2025-11-14T21:28Z [----] followers, [---] engagements
"CEOs of the biggest companies on the planet are shitposting at an unprecedented rate We have entered a new world (See sundar pichai posting about polymarket odds for his own products next release) Soon. https://t.co/KoEA6KjnO8 Soon. https://t.co/KoEA6KjnO8"
X Link 2025-11-16T19:16Z [----] followers, [----] engagements
"Forecasting $NVO's EVOKE/EVOKE+ semaglutide in early-stage Alzheimer Disease trial results Open to discussion please share thoughts and comments. https://www.aditharun.com/p/semaglutide-in-early-stage-alzheimers https://www.aditharun.com/p/semaglutide-in-early-stage-alzheimers"
X Link 2025-11-17T14:06Z [----] followers, 25.2K engagements
"@blippd Sentiment is very low and AD trial takes many years. But how many patients can they actually capture As in patients who dont already rule in for MACE or T2DM or other indications. I think 20% may happen if readout is above that seen in anti amyloid ab drug performance"
X Link 2025-11-17T17:46Z [----] followers, [---] engagements
"Agreed Safety is key and we know glp1 side effect profile well this trial wont unlock new safety signals. At that point its up to patients and doctor shared decision making imo. I favor this especially because there is a high unmet need and we have no good option so semaglutide is better than nothing or incredibly expensive brain bleed risk anti amyloid drugs"
X Link 2025-11-17T21:42Z [----] followers, [---] engagements
"PH-HFpEF space heating up 🔥 - $tecx relaxin mimetic ph2 [----] - $tenx levosimendan (inodilator) ph3 [----] - $azn relaxin mimetic ph2 readout [----] (still waiting) - $mrk activin inhibitor ph2 positive (announced today) / already approved in PAH (group [--] PH) to initiate phase [--] There is a large unmet need for patients with PH including group [--] PH (PH associated with left heart disease). It is wonderful that by end of 2026/2027 we could have new treatment options for this disease Will Q4 $TECX Ph1b topline readout of long-acting relaxin (TX45) in pulmonary hypertension (PH) associated with HFrEF"
X Link 2025-11-18T16:50Z [----] followers, 21.9K engagements
"@ResearchPulse1 What can I read on manufacturing side to learn about small molecule and peptide synthesis costs logistics and procedure"
X Link 2025-11-18T17:04Z [----] followers, [--] engagements
"@ResearchPulse1 @researchitgg Oh I see. But who cares Adverse effects matter but so far its pretty similar. And the 73x only matters insofar as it influences cost of the whole system (making and distributing pill vs injectable)"
X Link 2025-11-18T17:52Z [----] followers, [--] engagements
"@ResearchPulse1 @researchitgg Oral small molecules scale faster. But oral glp1 peptide vs sq injection is the question and lly/nvo are ramping oral production and pill stockpiles. Distribution is easier too"
X Link 2025-11-18T18:31Z [----] followers, [--] engagements
"@avidresearch @DrSamuelBHume AZN still hasn't shared data but also hasn't killed their program afaik TECX and TENX no expected readouts till [----] So all is quiet till end of Q1 [----] except for AZN"
X Link 2025-11-18T23:43Z [----] followers, [---] engagements
"@plainyogurt21 $MRK is going after a population with more severe disease and RV dysfunction (if you accept that PVR is a good marker of RV dysfunction in HFpEF)"
X Link 2025-11-19T00:11Z [----] followers, [---] engagements
"In other words $fulc (600M mc) is trading at a little under 1/3 the value of $beam (2.1B mc) $FULC +18% to 52w high off $AGIO p3 P1b 20mg cohort data at ASH $FULC +18% to 52w high off $AGIO p3 P1b 20mg cohort data at ASH"
X Link 2025-11-19T17:06Z [----] followers, [----] engagements
"Who will generate successful phase 2/3 data for AI-designed antibody first Assume bindcraft coerced to make antibody designs only Nabla Chai Other / show me results BindCraft Nabla Chai Other / show me results BindCraft"
X Link 2025-11-20T16:19Z [----] followers, [---] engagements
"@3rdFloorCapital Yeah. Traded up AH and early session but now down. Maybe this is just mild rerating given jobs report and now there will be no cuts Prediction markets had 60% chance of rate cut a few days ago now its like 20% so maybe this is a function of that news idk"
X Link 2025-11-20T17:53Z [----] followers, [---] engagements
"$FULC is the most promising sickle cell disease drug currently on the market. It also has the least data available. I will articulate my reasoning here: Gene editing approaches require chemo conditioning requiring in patient admission dealing with chemo toxicities (pretty bad) high up front cost (millions of dollars). There is an ongoing question as to whether the median patient wants this. Scalability remains a concern. $CRSP and $VRTX make Casgevy (FDA approved) -- apheresis busulfan conditioning then infuse ex-vivo edited hematopoietic stem cells (cas9 double stranded break of bcl11a"
X Link 2025-11-20T20:57Z [----] followers, 16.4K engagements
"Lumina is an interesting concept but I don't think the scientific evidence for its efficacy exists. Also we need human testing data. I get not wanting to register and go down FDA path but some human testing data is warranted before selling. Thoughts below but TL;DR I think you actually create periodontal issues with the probiotic. Also caries scoring system is based on [--] years ago and vague/not validated/highly subjective. Lumina probiotics per their company website is a probiotic optimized through biodesign to rebalance your oral pH and inhibit the growth of acid-secreting bacteria to"
X Link 2025-11-21T20:19Z [----] followers, [----] engagements
"@gen0m1cs I see it the same way. Downside risk is real given the scientific backing this product stands on. And I suspect this is something where cumulative exposure will matter. This also makes it harder to test because we would need to observe outcomes over a longer period of time"
X Link 2025-11-21T21:18Z [----] followers, [--] engagements
"@princetongb Whats your abvx pt for buyout"
X Link 2025-11-23T19:08Z [----] followers, [---] engagements
"I was wrong semaglutide does not change clinical dementia rating score trajectories meaningfully compared to placebo in patients with mild dementia due to Alzheimer's disease. This is a setback for patients but something that I felt needed to be tested given the reasons I covered in the quoted post. It may be that we did not offer the intervention early enough in the disease course. By the time patients with Alzheimer's disease are symptomatic / have imaging findings the disease process has been in motion for at least a decade. Perhaps an intervention like GLP1s that dampen inflammation and"
X Link 2025-11-24T17:07Z [----] followers, 14.4K engagements
"From the $nvo stock perspective I anticipated a plus/minus 5-10% move on trial results as article in my substack post a week ago I leaned towards the trial hitting statistical significance in CDR-SB but was not overwhelmingly sure. Net-net I felt the risk-reward was there given the anticipated move and the way the market was thinking about this trial readout. Sized my position accordingly and was wrong. $NVO fell 10% immediately after announcment and is now down 6% - within my expectations. Totally fine with the result and happy with sizing and way I structured trade. http://www.aditharun.com"
X Link 2025-11-24T17:20Z [----] followers, [----] engagements
"The stage of disease matters and in the re-analysis of the RCTs of GLP1-RAs time to dementia was the endpoint that showed benefit [--]. Incidence of disease is different from progression of disease. GLP1 for risk factor control and anti-inflammation WAY before disease even manifests clinically may be very useful. Perhaps individuals who are homozygous for APOE4 may be poised to receive benefit from VERY early use of GLP1s [--] Nrgaard et al. Treatment with glucagon-like peptide-1 receptor agonists and incidence of dementia: Data from pooled double-blind randomized controlled trials and nationwide"
X Link 2025-11-24T17:23Z [----] followers, [---] engagements
Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing