AI is getting really good at tasks with verifiable rewards namely coding.
Computational biology feels like coding but it is not. Much of biology is about identifying and using symbolic causal representations of how things work, and hierarchical values guiding experiments, data, benchmarks.
Why do schizophrenia GWAS signals look so flat across the genome?
In our recent preprint, we explored why psychiatric disorders — and, more broadly, brain-related traits involving the central nervous system — appear to have unusual genetic architectures.
🧵1/n
Two racks of ribs on a smoky grill surrounded by snow
No such thing as bad grilling weather, only bad grilling clothing
Not shown: the large snowflakes that fell on the ribs
Agreed! Our family in Texas wouldn't believe us when we said we were happy to grill for them in 30-40 degree weather, and that is actually the best time to do it with the grill keeping us warm....
I really hate it when scientists keep saying that “we need to rebuild trust in science,” because it implies that scientists are to blame for the mistrust rather than the millions of dollars of dark money that have funded political attacks on science in order to advance a far right agenda.
This is functionally a robbery of New York City and New York State taxpayers, who contribute billions more to the federal government annually than we get back.
Stealing our money, blocking our bus lanes.
"In this review, we provide a comprehensive overview of the clinical features, disease course and management of AYA patients with MPN and, in doing so, highlight key characteristics that distinguish them from their older counterparts."
onlinelibrary.wiley.com/doi/10.1111/...
"Sweeteners can harm cognitive health equivalent to 1.6 years of ageing, study finds"
Or does it? Let's take a look at this "study"...
www.theguardian.com/food/2025/se...
Last week, I talked to a DC-based PhD student who told me she's now afraid to leave her house after dark. Not because of any crime threat, but because of all the federal agents swarming her neighborhood.
Shortcomings of silhouette in single-cell integration benchmarking - @uweohler.bsky.social @prauten.bsky.social @mdc-bimsb.bsky.social @mdc-berlin.bsky.social @humboldtuni.bsky.social go.nature.com/4fcQzZr
Well this is bad. Google is indexing ChatGPT conversations exposing sensitive user data
I tried a few quick searches. I found someone's chat where I can see their api key
I found some building their resume. Their name, email and phone numbers are exposed.
www.fastcompany.com/91376687/goo...
AlphaFold is great, but contrary to public opinion it has not completely solved the protein folding problem. Much work remains to be done.
clauswilke.substack.com/p/no-alphafo...
Glad I'm not the only one observing the irony of tech billionaires handing out science prizes while backing an admin that has dismantled American science. New level of #respect for Seth Rogan
Has anyone else been having issues accessing papers on biorxiv?
More in No, Other People:
New poll:
“America would be better off if more people worked in manufacturing.”
• 80% of Americans agree
• 20% disagree
“I would be better off if I worked in a factory.”
• 25% of Americans agree
• 73% disagree
• 2% currently work in a factory
Rubio on X: Last night, another 10 criminals from the MS-13 and Tren de Aragua Foreign Terrorist Organizations arrived in El Salvador. The alliance between @POTUS and President @nayibbukele has become an example for security and prosperity in our hemisphere.
Marco Rubio just announced 10 more people were sent to be imprisoned in CECOT. They were flown there from Gitmo to be imprisoned potentially for life based not on any crime for which they were convicted, but on unproven allegations of gang membership with no due process.
This is not lawful.
Booker: "This isn't just about Mr. Garcia. This is about every American understanding that this president is making an assault on the due process rights that are afforded to people in our country ... you cannot disappear people off American streets."
Making my inaugural Bluesky post to share my latest preprint! Here, we meta-analyzed #brain #transcriptomic data from seven #schizophrenia associated mouse mutant models in order to find convergence at the molecular level. Check out our findings here!
www.biorxiv.org/content/10.1...
🚀 New Preprint Alert! 🚀 With the Standards and Technology Working Group of the @humancellatlas.org, we present a new preprint guiding multimodal #singlecell data integration and human organ atlas generation! 📄
Read the preprint: www.biorxiv.org/content/10.1...
In our updated TraitGym preprint (w/ @gonzalobenegas.bsky.social & Gökcen Eraslan), we evaluate Evo 2 on regulatory variants associated with human traits. We see marked performance gains with scale on Mendelian traits, although still a bit behind alignment-based methods.
doi.org/10.1101/2025...
1/n
I really enjoyed my visit with @noorpratap.bsky.social to the Schmidt Center at the Broad Institute yesterday. Thanks especially to our host Tavor Baharav. The MIA seminar is a great resource, and you can find Noor's and my talks here:
youtu.be/cAJNqu50YkM?...
Just rewatched Arcane on Netflix, highly recommend if you haven't seen it, way better than one would think based on the 'source' material...
Improvements in LM (top) and EM (bottom) of our micro-sam model (finetuned) compared to the default SAM model.
After a long journey, Segment Anything for Microscopy is now published in Nature Methods! We significantly improve SAM for interactive and automatic segmentation in light and electron microscopy and build a user-friendly tool.
www.nature.com/articles/s41...
One of the reasons Horror is such an important genre right now — besides the fact that reality is a horror show — is that it’s a genre where people still regularly manage to make original and interesting films and increasingly that’s just not true about filmmaking in general
A month ago we @vevotherapeutics.bsky.social announced that we have generated the largest single-cell perturbation atlas in history, Tahoe-100M. Today, we announce that we will fully open-source Tahoe-100M in Feb, as part of a collaboration with NVidia health to train cell state models.
Out today in @naturegenet.bsky.social -- PERFF-seq! With @tsionabay.bsky.social , @ronanchaligne.bsky.social, Bob Stickels, Meril Takizawa, + Ansu Satpathy, we describe this new assay to study rare populations with programmable nucleic acid cytometry. 1/n
www.nature.com/articles/s41...
Very excited to announce that the single cell/nuc. RNA/ATAC/multi-ome resource from ENCODE4 is now officially public. This includes raw data, processed data, annotations and pseudobulk products. Covers many human & mouse tissues. 1/
www.encodeproject.org/single-cell/...
Just in time for Christmas, a new release of simpleaf! This is a big one (0.18.0). Now simpleaf has an atac sub-command to handle scATAC-seq preprocessing with alevin-fry-atac! 1/5
www.biorxiv.org/content/10.1...
A scientist's perspective on why this is a terrible article. A summary - a woman with a rare mutation that puts her at risk of a severe, untreatable, degenerative disease. She has dedicated her live to supporting the science to find a solution. Left there, it is a beautiful piece. 1/
(gift link)
Preprint + Xmas present from Maria Herrero, a gifted postdoc in my research group, working with colleagues in Søren Brunak's group - "Clinical + Biological Stratification in 121,560 Antidepressant Prescription Trajectories using Unsupervised Modelling and Clustering" www.medrxiv.org/content/10.1...
