Jason Sheltzer

Check out our new study in @ScienceMagazine, where we take on a 100-year-old debate: what’s the role of aneuploidy in cancer? We discovered that genetically removing extra chromosomes blocks cancer growth - a phenomenon we call “aneuploidy addiction”. science.org/doi/10.1126/sc…

@xinjin @perturbai_tx congrats Xin!

Today, I am thrilled to introduce @PerturbAI_Tx, a company I co-founded to rethink drug discovery. Biology is complex, and disease unfolds across genes, cell types, and tissues. We’re powering AI with the causal logic of human genetics: perturb.ai

Not sure how many people have noticed this: In the latest (Mar 4) release of 24 new research papers from Nature (top science journal), 8 came from the US and 8 came from China. nature.com/nature/researc… Among the 29 new research papers in Cell (top journal in biomedical sciences) released this year, 8 came from the US while 10 came from China. cell.com/cell/newarticl… Meanwhile US labs are suffering from the aftermath of the NIH/NSF funding disruptions - including drastically reduced funding rates and as a result diminishing lab sizes and competitiveness in #STEM research.

We think that the chromosome engineering techniques that we’ve developed through this work will be broadly useful for creating isogenic models of other complex rearrangements associated with developmental syndromes. If you’re interested in collaborating, please get in touch!

This project was a fantastic collaboration with the Pinter Lab at UConn and the Davoli Lab at NYU. We’re grateful to @chromosome8p for supporting this work and for the individuals who provided the samples used in this analysis. genome.cshlp.org/content/36/3/5…

Thrilled to share my lab’s new paper out in @genomeresearch where we use chromosome engineering to deconstruct 8p syndrome - a rare developmental condition caused by inversions, duplications, and deletions on chromosome 8p.

@ruth_hook_ two major unanswered questions IMO: (1) even if most aneuploid cells get expelled to the trophectoderm, does the embryo retain some degree of mosaicism? (2) is the degree of mosaic aneuploidy sufficient to result in problems with health or development?

this paper came out almost five years ago and imo it should have been national news based on how many older couples are working with such a low n of embryos to begin with tl/dr: (~oversimplified, please read their work!) early aneuploidy is normal and development autocorrects

My lab at @StanfordCancer is hiring a computational biologist to support our efforts studying alterations in tumor genomes. This is a great opportunity for individuals looking to gain additional research experience before applying to PhD programs. More info below - please share!

@whitefishlab @ScienceMagazine some of us are still trying to use it for that purpose!

@JSheltzer @ScienceMagazine This is crazy. Thanks for sharing this. Despite it all, twitter/x (whatever) is still a good way to disseminate science.

AI is cool and all... but a new paper in @ScienceMagazine kind of figured out the origin of life? The paper reports the discovery of a simple 45-nucleotide RNA molecule that can perfectly copy itself.

Amazing paper from @PhilHolliger science.org/doi/10.1126/sc…

And Multivac said, "let there be GGC AGG GCA GCG CAG UCG GAC AUU GAU AAC GGA AUC GAG CCC GCC", and there was life!

@boxiangliu @joans Thanks Boxiang!

@JSheltzer @joans Thank you. This will go into my lecture material!

Let’s play a little game. Let’s say that you’re the CSO at a cancer pharma company, and you have to choose a target to go after. Here’s a gene – high expression is associated with poor prognosis in brain cancer. Looks like a good candidate for an inhibitor right?

Princeton neuroscientist Sam Wang is running to represent New Jersey in the House of Representatives! It goes without saying - Congress needs more scientists.

@BenSiranosian can you give an example of a key bioinformatics decision where an LLM would provide the wrong answer?

@JSheltzer I’m giving people an LLM-encouraged assignment. It should be quick with agentic coding but still requires key bioinformatics decisions. Also ask how they use ai in their workflow and what those decisions and justifications are.

@thelunglorax Part of the goal is to avoid doing 40+ first round interviews and asking a question live!

@JSheltzer Curious to hear what you land on! You could always have them demonstrate ability to correctly pseudocode the problem in interview. Maybe worth intentionally designing the challenge around ability to implement AI tools knowledgeably + ability to sanity check the output though?