Dropping my son off at ultimate frisbee summer camp this week (why yes, that’s a thing!) and realizing this may be the most bay area thing we’ve ever done
English
Vineeta Agarwala
3.9K posts
Vineeta Agarwala
@vintweeta
gp @a16z bio | physician, adjunct clinical prof @stanfordmed | prev @gvteam, product @flatironhealth, mdphd @broadinstitute | out-tweeting @paraga @littleansh
Palo Alto, CA Katılım Mayıs 2011
1.3K Takip Edilen23.3K Takipçiler
Legit felt like I might cry while reading Charlotte’s Web to my kids this weekend
I had forgotten what a truly beautiful piece of writing this book is 🩶
“After all, what’s a life anyway? We’re born, we live a little while, we die…By helping you, perhaps I was trying to lift up my life a trifle. Heaven knows anyone’s life can stand a little of that.”
English
And boy did $LLY go shopping…! 🛒
The good news for biotechs is that the company is also a great partner, genuine supporter of the startup ecosystem, and resourced with both the $$ and the people to take great science (even if still high risk) all the way forward through expensive clinical development
Kyle LaHucik@ky_lahucik
Lilly inked 25% of biopharma industry's acquisitions in 1H26 W/tirzepatide’s patent cliff far away, Lilly can “go shopping in aisles of the supermarket that are a little less populated,” said Jacob Van Naarden, head of Lilly BD, in interview w/ @endpts endpoints.news/eli-lilly-domi…
English
Vineeta Agarwala retweetledi
Biomni was early to a category that now feels obvious
today, @phylo_bio feels prescient again with its model-agnostic approach, as frontier model access gets harder for bio use cases and frontier labs move closer to developing drugs themselves
biopharma companies will increasingly need leading bio-native agentic capabilities that aren’t locked to any single model provider, and can fully leverage open-source models
try Biomni Lab here: biomni.phylo.bio
kudos @KexinHuang5, @YuanhaoQ, @jure, and team on the Science publication!
Kexin Huang@KexinHuang5
Today, we're excited to share that Biomni is published in @ScienceMagazine. Biomedical research is still fragmented, manual, and difficult to scale. In this work, we introduce Biomni - the first general-purpose biomedical AI agent with an integrated biology environment that can reason, plan, and execute end-to-end scientific workflows. We show that, with the right environment and harness, AI can automate large-scale omics analyses, orchestrate laboratory robotics, optimize molecular properties, and even train new AI models for biology. We also introduce a reinforcement learning recipe for continually improving biomedical AI agents, enabling open-source models to achieve frontier-level performance. It's surreal to look back. We started the Biomni project in early 2024, when agentic AI was still nascent. It is exciting to see tens of thousands of biologists collaborating with agents every day to accelerate science. Try Biomni: biomni.phylo.bio Read more: science.org/doi/10.1126/sc… This work is not possible without this truly inter-disciplinary team: @serena2z @hcwww_ @YuanhaoQ Minta Lu, Ryan Li, @yusufroohani Lin Qiu @shiyi_c98 Gavin Junze Di @rickwierenga @kavi_deniz Sherry @TianweiShe Shruti Jennefer Xin Zhou @MWheelerMD Jon Bernstein @MengdiWang10 @PengHeAtlas @zhou_jingtian @SnyderShot @lecong Aviv Regev @jure @StanfordAILab @genentech @phylo_bio @arcinstitute @UW @berkeley_ai @RetroBio_ @tamarindbio @Princeton @UCSF
English
Had missed this detail — @Humana was selected by CMS as the ‘central processor’ for the Medicare GLP-1 Bridge
Makes sense given Humana already administers prescription drug access for the LI NET program
cms.gov/medicare/cover…
Vineeta Agarwala@vintweeta
The details of this @CMSGov "Medicare GLP-1 Bridge" (which runs entirely *outside* the part D payment flow) are actually kind of interesting Believe this is the first time CMS has set up this kind of price and access bridge to drive rapid access to a specific prescription med class? If there are prior examples would love to learn... Starting July 1, 2026 for 18 months, CMS will enable access to Foundayo, Wegovy, and Zepbound at a $50 co-pay for patients who meet clinical criteria, while manufacturers will receive $245/mo net pricing (this is regardless of the exact product - so clearly Lilly/Novo will prefer to sell orals) CMS is setting up its own, separate central processing for prior auth, claims adjudication, and pharmacy payment for these meds‼️- presumably to guarantee that patients won't fall into long and unnecessary prior auth / UM traps with their plans cms.gov/medicare/cover…
English
Vineeta Agarwala retweetledi
One of my white whales at @EvidenceOpen is for us to do a better job conveying the level (un)certainty inherent in the set of refs we’ve retrieved for a given answer. A step in that direction is rolling out in product today!
Musings on the problem and how we are approached it here:
openevidence.com/blog/introduci…
English
Vineeta Agarwala retweetledi
Great to see the MedScribe benchmark we designed with @ValsAI and Harvard used to judge current SOTA models.
Today, Protege holds over 2,000,000 visits worth of medical audio, constituting over 200,000 total audio hours, along with 1,000,000,000's of medical notes that correspond to hundreds of millions of SOAP notes.
For this benchmark and from this corpus, we curated medical records and SOAP notes and test the model's ability to generate accurate SOAP notes from a medical transcript.
Most importantly.... none of the benchmark cases have ever appeared in any foundation model training!
Getting this right is huge for any medical AI application. A SOAP (Subjective, Objective, Assessment, Plan) note is a key endpoint in care delivery frameworks, where a doctor writes the care plan and assessment for an office visit.
We'll continue to see models improve on real-world tasks like these so long as they have access to the real-world data that accurately reflects real-world conditions.
Alexandr Wang@alexandr_wang
Muse Spark 1.1 is SOTA on Harvey's Legal Bench, TaxEval, and MedScribe. It's cool to see that our model outperforms even Fable in a few areas :)
English
Vineeta Agarwala retweetledi
@vintweeta @EvidenceOpen Thank you! More details in a blog post coming soon.
English
Kudos @IAmSamFin and @EvidenceOpen team!
"EvidenceGrade surfaces, quantifies, grades and visualizes the quality of the evidence behind OpenEvidence answers...
...The model is trained to weigh study design strength, consistency and precision across sources, and how directly the evidence applies to the question—mirroring how an expert methodologist would appraise a body of evidence"
fiercehealthcare.com/ai-and-machine…
English
Vineeta Agarwala retweetledi
Today, we're excited to share that Biomni is published in @ScienceMagazine.
Biomedical research is still fragmented, manual, and difficult to scale. In this work, we introduce Biomni - the first general-purpose biomedical AI agent with an integrated biology environment that can reason, plan, and execute end-to-end scientific workflows.
We show that, with the right environment and harness, AI can automate large-scale omics analyses, orchestrate laboratory robotics, optimize molecular properties, and even train new AI models for biology.
We also introduce a reinforcement learning recipe for continually improving biomedical AI agents, enabling open-source models to achieve frontier-level performance.
It's surreal to look back. We started the Biomni project in early 2024, when agentic AI was still nascent. It is exciting to see tens of thousands of biologists collaborating with agents every day to accelerate science.
Try Biomni: biomni.phylo.bio
Read more: science.org/doi/10.1126/sc…
This work is not possible without this truly inter-disciplinary team: @serena2z @hcwww_ @YuanhaoQ Minta Lu, Ryan Li, @yusufroohani Lin Qiu @shiyi_c98 Gavin Junze Di @rickwierenga @kavi_deniz Sherry @TianweiShe Shruti Jennefer Xin Zhou @MWheelerMD Jon Bernstein @MengdiWang10 @PengHeAtlas @zhou_jingtian @SnyderShot @lecong Aviv Regev @jure
@StanfordAILab @genentech @phylo_bio @arcinstitute @UW @berkeley_ai @RetroBio_ @tamarindbio @Princeton @UCSF
English
CHMP just opened phased review of @RevMedicines daraxonrasib before the full MAA is even in — a mechanism EU has used sparingly...
"A phased review aims to accelerate the assessment of a medicine by evaluating the data in phases as they become available, ahead of the submission of a full marketing authorization application"
ir.revmed.com/news-releases/…
English
Thrilled to be promoted to Full Professor and honored to be the inaugural recipient of the Hibbert Professorship @Penn! A huge thank you to Alister and Sharon Hibbert for endowing this professorship and supporting our work on next-gen delivery technologies for genetic medicines.
Penn Engineering@PennEngineers
Penn Engineering is pleased to announce the establishment of the Hibbert Professor, a new scholarly professorship to be held by inaugural recipient Mike Mitchell (@MJMitchell_Lab), an internationally recognized leader in lipid nanoparticle technologies for mRNA and genetic medicines. Made possible through the generosity of Alister M. Hibbert and Sharon N. Hibbert, dedicated supporters of Penn Engineering, the professorship reflects a shared commitment to advancing education and research while supporting exceptional faculty for generations to come. @pennbioeng bit.ly/4vlVNsd
English
This is a post for the healthcare acronym-loving wonks who track CMS VBC programs (iykyk) and know how hard achievements like this are...
This team is actually bending the US healthcare cost curve in Medicare -- let's go @PearlHealth_!
🟣 Pearl equips providers with technology to deliver better care at scale, while also taking on complex financial risk in CMS shared savings programs
🟣 Pearl manages $3.6B in annualized medical spend across >250K Medicare beneficiaries in 40+ states
🟣 [now for the acronym salad] Pearl has helped providers succeed in every variation of VBC, from GPDC to ACO-REACH to MSSP and next year, LEAD
🟣 The company is projected to deliver $500M in gross savings and triple its patient base from 2024 through the end of 2026
This is already a rare constellation of outcomes in this category, especially when scaling entirely through AI + technology (rather than workforce expansion). But wait there's more...
🟣 Pearl reached profitability in 2025
We led the series A back in 2021, and have been proud ever since for @a16z to continue backing this team
prnewswire.com/news-releases/…
English
Messi giving millennials everywhere so much life right now
Excellence is a beautiful thing to behold !!
FOX Sports@FOXSports
CLUTCH FROM THE GOAT. Lionel Messi scores his 21st career FIFA World Cup goal 🐐
English
Congrats to the Myricx and @Novartis teams!
The demand for novel ADC payloads is indeed real — up to $1.5b for a novel NMTi payload and preclinical programs in large solid tumor markets
“Preclinical data suggest this novel NMTi payload may have broad activity across solid tumors, including TOPO-1 resistant models, and may enable more effective use of ADCs in settings where existing payload classes have limitations.”
novartis.com/news/media-rel…
Vineeta Agarwala@vintweeta
Why we’re excited about ADCs with dual payloads and novel payloads! And RLTs against ADC-validated targets. Hit the same tumor targets, but with different cell killing MoAs or combination payloads. This is “Chemo 2.0”
English
Wow, largest acquisition for @VertexPharma ever, and the company now operates *commercially* in 5 TAs
CF (Trikafta, Alyftrek; built in-house)
Heme (CASGEVY via @CRISPRTX partnership)
Pain (JOURNAVX; built in-house)
Renal (povetacicept; via Alpine acquisition)
Endocrine (Palsonify via @Crinetics acquisition)
Until 2023, 100% of the company's revenues were in CF!
Endpoints News@endpts
🚨 Breaking -- Vertex makes its biggest deal ever: a $10B takeover of Crinetics. Via @ky_lahucik endpoints.news/vertex-makes-1…
English
Cool! It's always seemed like UCB + UCSF could drive even more world class collaboration across AI, computational science, and biomedicine. Congrats @yun_s_song @fraser_lab
Yun S. Song@yun_s_song
I am thrilled to share that UC Berkeley and UCSF have launched a joint initiative in Computational Biomedicine! cdss.berkeley.edu/news/uc-berkel… We will soon be recruiting new faculty and postdoctoral fellows. Please repost to help spread the word.
English
Vineeta Agarwala retweetledi
3 ai product roles at @a16z-backed infinite healthcare startups:
1. counsel health [series a] — sr product manager, consumer
2. prosper ai [series a] — ai product manager
3. abridge [series e] — product lead, ai/ml
dm for intros
@CounselHealth @ProsperAI_HQ @AbridgeHQ
English
Strong co-sign. So well said.
“We don’t need the government to be fast nearly as much as we need it to be dated. A yes in twelve months is a project you can finance. A no in twelve months is a write-off, and write-offs we can live with. What capital can’t price is the undated ‘maybe,’ the review that runs two years or seven depending on litigation, staffing, and luck.
…The clock forces a decision, but it doesn’t force a yes. A legitimate system can say yes, say no, and explain why. What it can’t say is wait forever.”
English
Nice! @boltz_bio is on a roll, and true to the company’s open source roots, remains focused on putting access to the latest Boltz models in the hands of every scientist, all across entire pharma R&D orgs
(Note this is distinct from narrowly scoped discovery partnerships, or collaborations based on delivering binders against a select few targets — such partnerships have their place too!)
Core belief: AI models are best used by scientists inside an org who have intimate knowledge of the ‘hurdle’ to be solved for a given drug program, have access to all prior screening campaigns, are owning all downstream experimental data, and are seeing molecules advance against multiple targets side by side to prioritize resources / effort (and yes, token spend…) across programs
Congrats @GabriCorso, @boltz_bio, and @GSK teams!
Gabriele Corso@GabriCorso
Today, we are excited to announce a major partnership with @GSK to deploy the latest Boltz models across GSK’s research organization!
English