Sean Simmons

According to Chapter 18 of the U.S. Code, attempting to access classified documents without proper security clearance is a federal felony punishable by up to 10 years in prison. The rule of law should apply to everyone.

A month ago we @vevo_ai 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 @NVIDIAHealth to train cell state models.

Our GWAS meta-analysis on the genomic architecture of 40 circulating cytokines in 75,000 individuals is now out @CommsBio 🧬❗️ A short thread about the topline findings🧵

@StevenSalzberg1 I mean, at least based on a quick glance, all it would take is a few missed doublets or misannotated cells to have leakage (not sure if it is enough to change the results?)? Or maybe even from ambient RNA?

... the real question here should be, where is the leakage that allowed the model to predict so well? I think the training data contained the answers, but it's so complex the authors didn't realize it

Is this real, or is it excessive AI hype? New paper claims that an AI foundation model achieves experimental-level accuracy in predicting gene expression even in previously unseen cell types." I'll read the paper, but don't believe it for a second nature.com/articles/s4158…

Excited to share the final chapter my PhD research on @medrxivpreprint! eQTLs only explain 30% of GWAS loci and 11% of trait heritability. Here we investigate if epigenetic QTL can provide new insights at the remaining GWAS loci. medrxiv.org/content/10.110…

scRNA-seq data sets exploding in number and size - check out scanpy & anndata for >1b cells: new experimental update includes APIs for scaling with dask from anndata, integrated with lots of scanpy and rapids-singlecell functions. gist.github.com/ilan-gold/9880…

Thrilled to share our review on evolution of #MPN from normal #HSC, co-authored with @HormozLab & @bloodgenes & out as early view @Haematologica Hope its a good resource, we enjoyed putting to together, feedback welcome! #MPNsm Early View | Haematologica haematologica.org/article/view/1…

7 million pairs! A great resource for TCR-antigen interaction. TRAIT: A Comprehensive Database for T-cell Receptor-Antigen Interactions biorxiv.org/content/10.110…

1/ Introducing SCimilarity, a new foundation model to explore scRNAseq data across tissues and diseases! It learns a common measure of cell similarity by training a deep metric learning model on millions of cells from various human tissues and conditions. nature.com/articles/s4158…

Incredibly excited to see this brilliant work from @gheimberg #AvivRegev out this week. Years in the making, Graham thought of this before foundation models became cool :) A foundation model for searching & finding the closest cell state, and experimentally validated!

Not that it matters with a GOP trifecta, but Puerto Rico just decisively voted for statehood *again* with an absolute majority in a three-option race. Puerto Rico has more people than 22 states, all of them are US citizens, and yet it's completely left out to dry.

I am very skeptical of the AI/ML/computational methods to predict protein-protein interactions. Prove me wrong. I present a challenge for anyone who claims they can predict protein-protein interactions. 1/n

🧬 "Two subtle problems with over-representation analysis" highlights critical limitations in popular enrichment tools, e.g.: background reduction & false discovery rates. These challenges impact reliability of results in #RNAseq data analysis. doi.org/10.1093/bioadv… #GeneSetEnrichment @mdziemann @Bora_Anusuiya08

I’m excited to share an R package that we’ve created for mapping scRNA-seq data across normal and malignant hematopoiesis! We've used this to identify hematopoietic stem cells from healthy tissues and map leukemia cells from hundreds of patients. 1/6 github.com/andygxzeng/Bon…

Never ceases to amaze me how many folks who’ve heard RATM are in Paul Ryan mode, having literally ZERO understanding of anything that band was about and even less understanding where any of us might stand on contemporary issues. Recently was talking to a couple at a restaurant who were big fans of “Killing In The Name”. The nice lady said, “I love that song. It helped me rage against my parents and later against the jab!” I said, “Ma’am that song is about racist cops who often behave like the Ku Klux Klan in service of historical white supremacy and are boot licking lackeys and thugs of the racist capitalist ruling class.” She sat there chewing and blinking, chewing and blinking.

I’m not a puppet Elon.. I’m a daughter of two immigrant parents that had to work their ass off to provide for me! I’m a product of welfare, I’m a product of section 8, I’m a product of poverty and I’m a product of what happens when the system is set up against you….But you don’t know nothing about that. You don’t know not one thing about the American struggle…. PS fix my algorithm

Rashida Tlaib is doing GOTV in Michigan. The Uncommitted movement are up on the air with anti-Trump ads. These folks deserved so much better from the party. But in spite of it all, they know how much is at stake and they're throwing in to fight against the fascist movement.

Trying to wrap my head around these figures. Speechless isn’t the word.

Ajax @AjaxThx is now a clinical-stage company! First patient with myelofibrosis #MPNsm has been dosed w/ AJ1-11095, the 1st type II JAK2 inhibitor to reach clinical trials. "FPFV" is always such an exciting milestone. Grateful to patients & investigators. ajaxtherapeutics.com/ajax-therapeut…

Congrats to our friends at the @broadinstitute on the launch of Cell Annotation Service (CAS). CAS is a ML tool trained on over 87M+ cells from the CZ #CellxGene corpus to make identifying cell types and states in #scRNAseq studies more efficient