PhD candidate @YaleCBB in the lab of @hattaca | interested in single-cell genomics, drug discovery, and AI | former data analyst @NUFeinbergMed | alum @WashuBME
Joined September 2023
- Tweets50
- Following264
- Followers42
- Likes548
Sam Fenske retweeted
It's amazing to me that 6 years later, this software is still widely used.
First it survived a generation of single-cell foundation models and now it's up against powerful coding agents.
We (@GalenXing, @_romain_lopez_) just released the first beta of scvi-tools (via PyPi). Please try it out and report any issues! We still have a lot of work to do on the back-end related to ease of creating new models, but the front-end API is stable!
scvi-tools.org/en/latest/ind…
Sam Fenske retweeted
Announcing today that Noetik has achieved the first near-term milestone in our landmark collaboration with @GSK.
Earlier this year, we delivered and deployed the OCTO-VC model inference and fine-tuning capabilities, meeting GSK's testing criteria. Because the models simulate gene expression, cell states and tumor-immune interactions, GSK researchers can generate data for hypothesis exploration instantly speeding their learning cycles.
Read our announcement and learn more about our work to close the gap between AI models and medicines that work for patients: businesswire.com/news/home/2…
Sam Fenske retweeted
A report card on how epigenetic ("biological") aging clocks respond to interventions purported to promote longevity @NatureMedicine
nature.com/articles/s41591-0…
nature.com/articles/s41591-0…
@prof_horvath @rv_sehgal
Sam Fenske retweeted
How does aging reshape a tissue that must continually rebuild itself hundreds of times?
Using spatial transcriptomics, we find that ovarian aging disrupts the coordination of cells across space and time before reproductive cycles stop.
#OvarianAging
🗞️Our June issue is live!✨Read about the healthspan and lifespan effects of time-restricted feeding in mice, insights from mathematical models of aging, how R-loops drive the SASP, & much more nature.com/nataging/volumes/…
Sam Fenske retweeted
It’s estimated that the Protein Data Bank (PDB) cost around $13B to create. Alphafold was only possible because of it. If we want ML to solve biology, we should be funding the creation of databases and the development of new assay technologies. ML is nothing without data.
Sam Fenske retweeted
Every cancer drug you prescribe today
was once an NIH grant nobody noticed.
Pharma celebrates launches.
Few notice the 20–40 year runway behind that.
Gleevec took 41 years from NIH-funded discovery of the Philadelphia chromosome → FDA approval.
Behind every “breakthrough” sits a graveyard of failed attempts that made it possible.
Cut that lineage today,
the next Gleevec doesn’t arrive in 2067.
Funded science matters.
If someone says NIH funded science is inefficient,
show them this.
- - - - -
Source: NIH RePORTER · FDA · via @Jori_health
- - - - -
Sam Fenske retweeted
Very excited to present our work in NOLA exploring epigenomics in thoracic aortic aneurysms! Please consider passing by our moderated poster on Saturday @samf_7927 @hattaca @RolandAssiMD #DustinLieu #GeorgeTellides #ChungLab #AHA25 @YaleVBTProgram @YaleIMed @yalectsurgery
Sam Fenske retweeted
Ask not why would you work in biology, but rather: why wouldn't you?
owlposting.com/p/ask-not-why…
openai/gemini gave this essay an A- for evocative imagery. claude gave it a C- for being emotionally manipulative. both are probably right. i feel a little sick re-reading it
Sam Fenske retweeted
A recent study, led by Catherine Gao, PhD (@CatGaoHow), explores how machine learning models like LSTM and XGBoost can analyze electronic health record data to forecast next-day extubation with high accuracy. The findings suggest machine learning could support, not replace, clinical judgment in ventilator liberation decisions.
📄 Read the full study: nature.com/articles/s41598-0…
Sam Fenske retweeted
Excited to share that we've received an @NIH R35 MIRA! This grant will support our work building tools to study how cell states change over space and time in disease. Grateful to our community, and especially the wonderful team I get to work with.
🚨 Our paper is out in Scientific Reports! Amazing collaboration with @alecpeltekian, @CatGaoHow, and Ankit Agrawal @NUFeinbergMed
nature.com/articles/s41598-0…
In failed extubation cases, the model would’ve advised against extubation 35% of the time, showing promise as a second opinion tool 💡. Top predictors from SHAP and ablation testing include plateau pressure, heart rate, PaCO2, aligning with clinical intuition.
1
1
69
Sam Fenske retweeted
Opportunistic A.I. interpretation of medical scans.
The list keeps getting longer of improved detection, unrelated to why the scan was obtained
Today; Aortic aneurysm from breast MRI
nature.com/articles/s41467-0…
My running list👇
Sam Fenske retweeted
What if LLMs could “read” & “write” biology? 🤔
Introducing C2S‑Scale—a @Yale + @GoogleAI @GoogleDeepMind collab: we scaled LLMs (up to 27 B!) to analyze & generate single‑cell insights by turning transcriptomes into text 🧬➡️📝
🔗 Blog: research.google/blog/teachin…
🔗 Preprint: biorxiv.org/content/10.1101/…
#SingleCell #AI #LLM
Sam Fenske retweeted
Regeneron CEO Len Schleifer on @CNBC: The biotech/pharma sector should be seen as the "health defense industry" vital for global wellbeing. To sustain innovation, it needs proper support.
cnb.cx/4iQSkfV
