Sign in

Anshul Kundaje

@anshulkundaje.bsky.social
9.5K followers 4.5K following 966 posts

Genomics, Machine Learning, Statistics, Big Data and Football (Soccer, GGMU)

PostsRepliesMedia
Anshul Kundaje @anshulkundaje.bsky.social · 18/05/2026
Oh I'm surprised it's not released. I thought it was public. I'll ask the student to release it ASAP. It's pretty early work btw. Lots to do for stronger evals and baselines.
020
Reposted by Anshul Kundaje
Selin Jessa @selinjessa.com · 14/04/2026
We are so excited to see our work out in @nature.com! We present a multi-omic single-cell atlas of 12 organs in human fetal development, explore the enhancer landscape, use deep learning to infer rules of transcription factor activity, and interpret non-coding variants in complex traits: #GeneReg 🧬🖥️
14220
Reposted by Anshul Kundaje
Anthony Mathelier @amathelier.bsky.social · 02/12/2025
Proud of JASPAR 2026 release with major expansion of PFM profiles and a new collection of deep learning models interpreted to extract learnt motifs! Work led by @ievarau.bsky.social and @damlaob.bsky.social w/ @anshulkundaje.bsky.social @wywywa.bsky.social @flowerwhatelse.bsky.social labs & more
053
Reposted by Anshul Kundaje
Zeitlinger Lab @zeitlingerlab.bsky.social · 19/11/2025
We are pleased to announce a new preprint by @mlweilert.bsky.social: “Widespread low-affinity motifs enhance chromatin accessibility and regulatory potential in mESCs” (www.biorxiv.org/content/10.1...). See summary and longer recap below: (TLDR; low-affinity motifs matter as pioneers!)
biorxiv.org
Widespread low-affinity motifs enhance chromatin accessibility and regulatory potential in mESCs
Low-affinity transcription factor (TF) motifs are an important element of the cis-regulatory code, yet they are notoriously difficult to map and mechanistically incompletely understood, limiting our a...
16928
Reposted by Anshul Kundaje
Max Planck Institute for Molecular Genetics @molgen.mpg.de · 07/11/2025
🧬📢 Looking forward to our next speaker in our lineup for the #Embryo2026 meeting, Anshul Kundaje from Stanford University. He will speak on the topic of "Deep learning the regulatory syntax of fetal development and the genetic basis of developmental disorders".
1142
Anshul Kundaje @anshulkundaje.bsky.social · 18/10/2025
Thanks. Happy to help if you run into any issues. If the goal is differential signal prediction, would need a slightly different fine tuning set up for optimal performance
010
Anshul Kundaje @anshulkundaje.bsky.social · 16/10/2025
Should have used this!
020
Anshul Kundaje @anshulkundaje.bsky.social · 15/10/2025
And just to be clear, I have no specific comments on the other primary models implemented in the CRESTED package including the ones they developed. They look good from the paper. I'm simply stating that the "Dilated CNN" implementation in the package is NOT ChromBPNet. 4/4
110
Reposted by Anshul Kundaje
Oded Rechavi @odedrechavi.bsky.social · 15/10/2025
Here's the link to the system, try it! qedscience.com @qedscience.bsky.social
qedscience.com
q.e.d Science
Critical Thinking AI for constructive criticism and science evaluation
1720467
Reposted by Anshul Kundaje
Oded Rechavi @odedrechavi.bsky.social · 15/10/2025
BIG ANNOUNCEMENT📣: I haven’t been this excited to be part of something new in 15 years… Thrilled to reveal the passion project I’ve been working on for the past year and a half!🙀🥳 (thread 👇)
56494184
Reposted by Anshul Kundaje
Luca Pinello @lucapinello.bsky.social · 15/10/2025
🧬Excited to be at #ASHG2025! My lab is presenting 5 studies this week. Join us for talks and posters on #SingleCell #CRISPR, #DeepLearning & more! Thread below with all presentation details 👇
2253
Reposted by Anshul Kundaje
Marisa Kabas @marisakabas.bsky.social · 15/10/2025
NEW — I dug into the story of ICE abducting a 13-year-old boy in Massachusetts and moving him 500 miles across state lines without notifying his mother, local cop's complicity, and how the federal government's justification for such cruelty is already falling apart:
thehandbasket.co
ICE took a 13-year-old they said had a gun. Local cops say he didn’t.
Now he's detained 500 miles away from his Massachusetts home.
12375523215
Anshul Kundaje @anshulkundaje.bsky.social · 15/10/2025
Thanks
000
Anshul Kundaje @anshulkundaje.bsky.social · 15/10/2025
Came across this interesting paper www.nature.com/articles/s41... that compares their model to a "Dilated CNN" model from the CRESTED package that is supposed to "ChromBPNet inspired". Performance of the dilated CNN is miserable. This is categorically NOT a ChromBPNet model 3/3
nature.com
IceQream: Quantitative chromosome accessibility analysis using physical TF models - Nature Communications
Cis regulatory elements endow genomes with sequence-encoded logic to drive cellular differentiation. Here, the authors introduce a biophysically principled sequence model that characterises complex TF...
210
Anshul Kundaje @anshulkundaje.bsky.social · 15/10/2025
Please don't refer to models from other packages as ChromBPNet. There is much more to ChromBPNet than the architecture. The training procedure, bias correction & data sampling strategy has a huge impact on performance. 2/
100
Anshul Kundaje @anshulkundaje.bsky.social · 15/10/2025
Friendly reminder that there is only one official release of ChromBPNet at our lab GitHub github.com/kundajelab/chr… There are other versions floating around eg. in the CRESTED package that are NOT anywhere close to the official implementation. 1/
github.com
11152
Reposted by Anshul Kundaje
Camille Moore @camille-moore.bsky.social · 02/10/2025
Today I am so pleased to present our work on how chromatin remodelers affect mesoscale chromatin organization. www.science.org/doi/10.1126/...
science.org
ATP-dependent remodeling of chromatin condensates reveals distinct mesoscale outcomes
Adenosine triphosphate (ATP)–dependent chromatin remodeling enzymes mobilize nucleosomes, but how such mobilization affects chromatin condensation is unclear. We investigate effects of two major remod...
48835
Reposted by Anshul Kundaje
ralucagordan.bsky.social @ralucagordan.bsky.social · 03/10/2025
I'm very pleased to announce the official publication of our lab's paper "DNA mutagenesis driven by transcription factor competition with mismatch repair" in today's issue of Cell! www.cell.com/cell/abstrac...
cell.com
DNA mutagenesis driven by transcription factor competition with mismatch repair
Competition between transcription factors and mismatch repair machinery drives localized hypermutation at regulatory elements, with implications for cancer and genome evolution.
26617
Reposted by Anshul Kundaje
Marisa Kabas @marisakabas.bsky.social · 30/09/2025
SCOOP — The White House instructed all federal agency leadership on Tuesday to send out a mandatory message to workers blaming the impending potential government shutdown on House Democrats, The Handbasket was first to learn. Absolutely no modifications to the email were permitted. My story:
thehandbasket.co
Trump mandates all federal agencies send email blaming Dems for potential gov’t shutdown
Workers call the partisan email "a vile slap in the face."
19548082178
Reposted by Anshul Kundaje
Tong Wang @twangmdphd.bsky.social · 29/09/2025
1/ Happy to share our preprint from the Greenleaf Lab: beCasKAS, our method to directly detect CRISPR base editor off-targets in primary cells. We additionally show how non-coding edits can be triaged for epigenetic dysregulation using deep learning. www.biorxiv.org/content/10.1...
1134
Reposted by Anshul Kundaje
Nobu Hamazaki @nobuhamazaki.bsky.social · 26/09/2025
New paper from my lab and @jshendure.bsky.social lab! Led by the brilliant @zukailiu.bsky.social and @cxqiu.bsky.social. We tackled how anterior and posterior progenitor cells cooperate to self-organize into an embryonic structure (termed AP-gastruloid). (1/n) www.biorxiv.org/content/10.1...
35420
Reposted by Anshul Kundaje
Daniel Bolnick @danielbolnick.bsky.social · 26/09/2025
The GRFP announcement from NSF cuts out an entire cohort of 2nd year students from consideration, without warning. This is so deeply unfair that it warrants a formal protest from the scientific community. If someone wants to work with me to craft an open letter and solicit signatures, LMK.
917286
Reposted by Anshul Kundaje
Eric Topol @erictopol.bsky.social · 24/09/2025
The projected rise in global cancer deaths 1990 —5.9 million 2023—10.4 million 2050—18.8 million is yet another reason we need to get primary prevention into high gear, which is now possible but not getting done www.thelancet.com/journals/lan...
25482145
Anshul Kundaje @anshulkundaje.bsky.social · 22/09/2025
And I think this is just the tip of the iceberg. There will be many ways to push this further. Really huge breakthrough IMO from my first reading. Can't wait to use this & build on it. Huge congrats to the team! 3/3
160
Anshul Kundaje @anshulkundaje.bsky.social · 22/09/2025
Also an absolute showcase for what self supervised learning can do when done right. IMO, this is the first paper that really shows the power of self supervised genomic language models. 2/
181
Anshul Kundaje @anshulkundaje.bsky.social · 22/09/2025
This is truly an incredible breakthrough IMO. Really exemplifies what you get when deep domain expertise (popgen/evolution/disease genetics in this case) fuses with cleverly crafted ML. What u get r sleek, well thought out architectures that absolutely destroy the behemoths. Wow!! 1/
36414
Reposted by Anshul Kundaje
Yun S. Song @yun-s-song.bsky.social · 22/09/2025
We are excited to share GPN-Star, a cost-effective, biologically grounded genomic language modeling framework that achieves state-of-the-art performance across a wide range of variant effect prediction tasks relevant to human genetics. www.biorxiv.org/content/10.1... (1/n)
417690
Reposted by Anshul Kundaje
Yuri Pritykin @yuripritykin.bsky.social · 19/09/2025
Excited to share our new study: Single-cell multiomics reveals archetypal regulatory programs shared across CD4 and CD8 T cell subsets in viral infection bioRxiv: doi.org/10.1101/2025... 🧵 Key findings below 👇
doi.org
Single-cell multiomics reveals archetypal regulatory programs shared across CD4 and CD8 T cell subsets in viral infection
T cells protect against pathogens and tumors and can differentiate into various functionally distinct subsets. While each subset exhibits a characteristic epigenomic and transcriptional profile, essen...
1113
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
But lets be clear that glam journal publishing is a branding & marketing exercise. It is not in any way a step towards greater transparency, accountability & credibility. 12/12
030
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
... to not support the gate keeping system followed by glam journal publishing. But I also get that for them it is a very cheap (only 13K for gold open access) but effective branding tool. So I definitely do not blame them for taking advantage of it. 11/
120
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
So yes I would encourage them to release their models open weights, open source, with fabulous documentation & respond to community review. But also ... 10/
130
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
I would go further & claim that companies are especially are not bound by any of the constraints that masochistic academics impose on themselves by tying their futures to exploitative & wasteful gate keeping publishing systems. 9/
110
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
So kudos to Deepseek for being super transparent with their tech, models, documentation. But publishing this work in Nature or any other glam journal is entirely a branding and marketing exercise. Lets not pretend otherwise. 8/
120
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
Occasionally they do live upto the hype. But most of the time that's not the case. The whole gate keeping exercise of glam journals is based on vibes about what is "hot" at the moment. 7/
120
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
Lastly, I don't know how anyone can claim that publishing a paper in Nature is a way to dial back hype. It is literally a branding exercise. Papers published in glam journals are the most overhyped of all scientific literature. 6/
130
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
Secondly, there are plenty of studies published in journals including Nature that are full of flaws. So I don't understand how anyone can claim that journal peer review is reliable quality control. 5/
130
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
Firstly, open source, open weights models can be and are openly vetted, critiqued and developed on by many more than 3 anointed peer reviewers. That is true peer review by a massive community of folks with wide ranging expertise. 4/
140
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
Peer review by 3 random reviewers and gate keeping by an editorial team from a glam journal is utterly unnecessary. And the point being made here that this form of journal peer review adds some form of verification is utterly misleading. 3/
330
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
This commentary is skipping the key point that this LLM (DeepSeek-R1) is open source, open weights with transparent documentation and methods (released Jan 2025). That is a key difference relative to other closed source, closed weights LLMs (ChatGPT, Claude, Gemini etc). 2/
150
Anshul Kundaje @anshulkundaje.bsky.social · 19/09/2025
An editorial was published in Nature recently claiming that glam journal publication of LLMs (like DeepSeek-R1 this case) marks a step towards greater transparency, accountability & credibility www.nature.com/articles/d41.... I have thoughts ... 1/
t.co
https://www.nature.com/articles/d41586-025-02979-9
1190
Reposted by Anshul Kundaje
David Ochoa @d0choa.bsky.social · 18/09/2025
If you care about regulatory variants don’t miss the E2G extension of the @opentargets.org Platform. There are synergies that only a fully open science allows and this is a good one!
0111
Reposted by Anshul Kundaje
Jesse Engreitz @jengreitz.bsky.social · 18/09/2025
This extends code from the Open Targets Platform — huge thanks to @opentargets.org for this resource. I use it every day! A subset of the ENCODE-rE2G predictions are now available through the Open Targets Platform as well. Thanks @d0choa.bsky.social ! bsky.app/profile/open... 5/
151
Reposted by Anshul Kundaje
Jesse Engreitz @jengreitz.bsky.social · 18/09/2025
1. Search for your favorite GWAS variant and find predicted target genes and cell types (e.g. e2g.stanford.edu/variant/10_7...): 2/
E2G Browser - rs1250566
141
Reposted by Anshul Kundaje
Jesse Engreitz @jengreitz.bsky.social · 18/09/2025
Excited for a major milestone in our efforts to map enhancers and interpret variants in the human genome: The E2G Portal! e2g.stanford.edu This collates our predictions of enhancer-gene regulatory interactions across >1,600 cell types and tissues. Uses cases 👇 1/
39242
Reposted by Anshul Kundaje
Open Targets @opentargets.org · 18/09/2025
@riyavsinha.bsky.social @jengreitz.bsky.social sky.social @anshulkundaje.bsky.social y.social have made the ENCODE-rE2G data available to browse through the E2G portal, a custom-built extension of the Platform 👇 We plan to further integrate their data 👀 e2g.stanford.edu
e2g.stanford.edu
E2G
E2G is a tool based on the Open Targets Platform for predicting enhancer-gene interactions.
196
Reposted by Anshul Kundaje
Open Targets @opentargets.org · 18/09/2025
The Open Targets Platform autumn release is out! 🍂 We have ingested the full dataset of over 13 million enhancer-gene regulatory interactions in the human genome across 1,458 DNase-seq experiments covering 369 cell types and tissues from the ENCODE-rE2G model blog.opentargets.org/open-targets...
blog.opentargets.org
Open Targets Platform 25.09 has been released!
The September 25 release includes enhancer-gene data predicted by ENCODE-rE2G, molQTL credible sets for targets, new options for the variant structural viewer, and a number of data updates from GWAS C...
1229
Anshul Kundaje @anshulkundaje.bsky.social · 17/09/2025
New faculty position opening at @stanford-chemh.bsky.social . Especially interested in candidates with deep expertise and vision for the future of AI/ML/computational models of biomolecules. Please RT
0158
Reposted by Anshul Kundaje
Polly Fordyce @pollyfordyce.bsky.social · 17/09/2025
Please apply to our tenure-track faculty position at @stanford-chemh.bsky.social! We are searching for a new colleague working at the interface between computation and molecular sciences. See post below and pls forward widely! chemh.stanford.edu/opportunitie...
chemh.stanford.edu
Faculty Recruitment
02825
Reposted by Anshul Kundaje
Denny Carter @dennycarter.bsky.social · 17/09/2025
RFK Jr. is not well
The hearing has resumed.

Dr. Monarez catalogued a number of false statements that Kennedy made to her: “He said that C.D.C. employees were killing children and they don’t care. He said that C.D.C. employees were bought by the pharmaceutical industry. He said C.D.C. forced people to wear masks and social distance, like a dictatorship. And the one I think that hurt me the most was a particularly vivid phrase: He said during the Covid outbreak, C.D.C. told hospitals to turn away sick Covid patients until they had blue lips before allowing them to get treatment.”

She said he was “very, very upset” and very animated in their meetings.
18626127
Reposted by Anshul Kundaje
Yuri Pritykin @yuripritykin.bsky.social · 16/09/2025
Happy to share our new paper! We present high-resolution Micro-C maps across Drosophila embryogenesis, integrated with 149 ChIP-seq datasets. This systematic analysis enabled clustering of loops & boundaries into 4 distinct classes with distinct regulatory roles. www.cell.com/cell-genomic...
cell.com
3D chromatin structures precede genome activation in Drosophila embryogenesis
Dolsten et al. map the 3D chromatin architecture across early Drosophila development using high-resolution Micro-C, and, by integrative analysis with 149 public ChIP-seq datasets, identify four classe...
1113