iiinova — where your digital self keeps evolving

Reposted by Kashyap Chhatbar

Nute @nutedawn.bsky.social · 19/07/2026

Ok this is very good

91707381 0

Reposted by Kashyap Chhatbar

MwahahahahahadScientist @mads100tist.bsky.social · 18/11/2024

'Look, it's going to be an easy paper. Trust me. Six months. Tops'

A wooden statue of a monk being embraced by an skeleton in a "trust me bro" kind of attitude

29210 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 15/04/2026

Congratulations to you and the whole lab🍾🎉

010 0

Reposted by Kashyap Chhatbar

Duncan Sproul @sproullab.bsky.social · 10/04/2026

Very pleased to say that the paper version of our study examining the role of de novo DNMTs in cancer-associated hypomethylation is now out in @plos.org Genetics 🎉: doi.org/10.1371/jour... You can read more on the previous thread but here is a summary of the main updates! #epigenetics 🧵 1/9

75015 0

Reposted by Kashyap Chhatbar

FragileNucleosome @fnucleosome.bsky.social · 05/04/2026

Join us this Wednesday to hear about fascinating research from @kashyapchhatbar.bsky.social and @a-radzisheuskaya.bsky.social! If you're not already registered, click here to sign up: us06web.zoom.us/webinar/regi... you can also join the conversation at our discord: discord.com/invite/dXqT89r

0158 0

Reposted by Kashyap Chhatbar

Phil Ewels @ewels.bsky.social · 02/04/2026

Super excited to be launching two things today: #RustQC 🦀🧬 and rewrites.bio 🚀 I used AI to rewrite 15 RNA-seq QC tools into a single Rust binary (I've never written any Rust). It ended up being over 60x faster. Here's the story 🧵 seqeralabs.github.io/RustQC/

seqeralabs.github.io

Welcome to RustQC

Fast quality control tools for sequencing data, written in Rust.

38536 0

Reposted by Kashyap Chhatbar

Johan Duchêne @johanduchene.bsky.social · 30/04/2025

Antibodies don't work. NLRP3 is by far the most studied inflammasome — yet <30% of commercial antibodies were specific. Only 4 of 14 tested antibodies detected NLRP3 specifically. Antibodies must be validated before use. Full report by @ycharos.bsky.social : zenodo.org/records/1162...

811242 0

Reposted by Kashyap Chhatbar

sina b @sina.bio · 03/02/2026

If you work in AI for Science, take a moment to familiarize yourself with a common failure mode: paranormal citations (or paracites). Our paper describes them.

082 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 04/02/2026

Congratulations to the whole team! 🥳🎊👏🏼

110 0

Reposted by Kashyap Chhatbar

Aaron Quinlan (he/him) @aaronquinlan.bsky.social · 02/12/2025

We are thrilled to announce the first official release (v0.1.8) of #𝗯𝗲𝗱𝗱𝗲𝗿, the successor to one of our flagship tool, #𝗯𝗲𝗱𝘁𝗼𝗼𝗹𝘀! Based on ideas we conceived of long ago (!), this was achieved thanks to the dedication of Brent Pedersen. 1/n

quinlanlab.org

Intro to Bedder – The Quinlan Lab

5298151 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 26/11/2025

An amazing experience to be a part of this multi-disciplinary project within the Bird Lab! 🐦✨ We brought together molecular biology🧩⛓️, genomics🧬🧶, and explainable machine learning📈🖥️🤖 to understand mechanisms of SALL4 function. So proud of this team effort! 🤝🙌 Check out the full breakdown below 🧵👇

050 0

Reposted by Kashyap Chhatbar

Duncan Sproul @sproullab.bsky.social · 25/11/2025

🚨 Multiple #PhD opportunities for lab or computational applicants. Do you want to understand the molecular basis of #epigenetic dysfunction in human disease? Come work with our group at the fantastic, collaborative @uoe-igc.bsky.social in a beautiful, liveable city. 🧵 1/4 Please repost 🙏

11916 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 24/10/2025

This is very cool! Can't wait to read it in detail. Congratulations to the whole team.

000 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 24/10/2025

💡 The big picture: This framework shows how integrating predictive modeling + XAI can generate and test novel biological hypotheses using only standard, wild-type data. It’s a powerful way to untangle complex regulatory networks before running expensive perturbation experiments!

000 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 24/10/2025

🎯 Why this matters for research: This XAI approach does more than just explain-it predicts. Genes ranked by our SHAP values are far more likely to be true direct targets than genes ranked by simple ChIP-seq signal. It even helps predict the magnitude of the transcriptional change!

(B) Fraction of overlapping direct target genes with genes ranked by summed SHAP values of INTS11 and ZC3H4 (blue lines) compared to genes ranked by normalised ChIP-seq signal (grey lines). (C) Absolute pearson correlation between SHAP values from different algorithms and log2 fold change following acute depletion of INTS11 and ZC3H4. As a control comparison, normalised ChIP-seq was correlated with log2 fold change. Error is estimated by calculating correlation coefficients for all 5 splits and the error bars indicate the standard deviation of one-half.

100 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 24/10/2025

Reality check ✅ ZC3H4 (Restrictor) and INTS11 (Integrator) target many of the same genes-then move them in tandem (r=0.84). 📈 Mimoso et al. report ~50% overlap in Restrictor-sensitive mRNAs. Together, this argues for deep functional coupling in transcription control 🧬

(B) Overlap of direct targets of SET1A, ZC3H4 and INTS11. (C) Scatter plot of log2 fold changes following ZC3H4 degradation and INTS11 degradation for the overlapping genes.

100 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 24/10/2025

🤯 Key Finding 2: We found an unexpected link between the Restrictor (ZC3H4) and Integrator (INTS11) complexes! Our model showed that INTS11 target genes were also highly dependent on ZC3H4, and vice-versa. This points to major crosstalk! 🤝

SHAP values for chromatin-associated proteins profiled in Restrictor model and Integrator model. Absolute mean DeepSHAP values computed for (A) direct targets (INTS11 degron) and (C) direct targets (ZC3H4 degron) and random genes at promoter and gene body contexts. Random subsampling is used to calculate intra-category variability and the error bars indicate the standard deviation of one-half.

100 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 24/10/2025

🔥Key Finding 1: ZC3H4 acts beyond promoters. While its job at promoters is known, we revealed it also has a significant regulatory contribution within the gene body. Consistent with ZC3H4 recruiting PNUTS-PP1 (Erickson; Kelly et al.) which functions across gene bodies #Epigenetics #GeneRegulation

SHAP values for SET1A, H3K4me3 and ZC3H4. (A) Chromatin-associated proteins and histone marks relevant for this model. (B) Comparison of absolute mean DeepSHAP values computed for direct targets (ZC3H4 degron) and random genes at promoter and gene body contexts. Random subsampling is used to calculate intra-category variability and the error bars indicate the standard deviation of one-half.

110 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 24/10/2025

But is it just correlation? 🤔 We validated our model's insights using degron experiments (for rapid protein degradation). 💥Genes most affected by the protein's removal were exactly the ones our model (trained on unperturbed data!) had flagged as most important. This shows functional relevance!

SHAP values for co-occupying PRC1 (RING1B) and PRC2 (SUZ12) components and their catalytic marks. Chromatin-associated proteins and histone marks relevant for this model. Comparison of absolute mean SHAP values computed for direct targets (RING1B degron 2h) and random genes at promoter gene body contexts.

120 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 24/10/2025

How does it work? 🤖 We trained an AI on data from unperturbed cells to predict gene transcription (RNA Pol-II occupancy). The secret sauce? We fed it signals from both promoters and gene bodies, then used SHAP (explainable AI) to tell us exactly which proteins matter most, and where!

Overview of predictive modelling for RNA Pol-II occupancy. Schematic of the multilayer perceptron (MLP) model architecture used to predict RNA Pol-II occupancy based on chromatin-associated protein profiles. Input features include protein occupancy data from ChIP-seq experiments, and outputs represent predicted RNA Pol-II occupancy.

100 0

Kashyap Chhatbar @kashyapchhatbar.bsky.social · 05/02/2025

Excited to share my latest preprint! 🚀 We infer regulatory relationships between chromatin-associated proteins and transcriptional control.

111 0

Reposted by Kashyap Chhatbar

Jess Kelley @jesskelley.bsky.social · 26/11/2024

Very excited that our paper describing a critical role for the PNUTS-PP1 phosphatase complex in transcription pause release is up online! This was a massive team effort with @edimitrova.bsky.social and the rest of the @robklose.bsky.social lab www.cell.com/molecular-ce... A short thread: (1/8)

cell.com

The PNUTS phosphatase complex controls transcription pause release

Kelley et al. discover that the PNUTS phosphatase complex plays an essential role in gene transcription by controlling RNA polymerase II pause release. PNUTS achieves this through its TND, which recog...

26628 0

Reposted by Kashyap Chhatbar

Rob Klose @robklose.bsky.social · 25/11/2024

Looking forward to sharing a new publication from the lab tomorrow! Something a bit different from us this time around. Stay tuned….

2362 0

Reposted by Kashyap Chhatbar

Madhu Pai, MD, PhD @madhupai.bsky.social · 24/11/2024

Non-native English speakers need 50% more time to write a paper When they do, they face a 2.5 times higher chance of being rejected because of language journals.plos.org/plosbiology/...

24642290 0

Reposted by Kashyap Chhatbar

Surfin' NuRD @surfinnurd.bsky.social · 23/11/2024

Loved being back in Edinburgh. Such a beautiful city, such great science. Great to catch up with Adrian, @cathyabbott.bsky.social @cellysally.bsky.social @hannahlong.bsky.social @kashyapchhatbar.bsky.social @r-pantier.bsky.social et al. Thanks Cath+Adrian for the hospitality and Congrats Sara!

3134 0