A large-scale study identified 380 regulatory SNVs linked to cancer risk, impacting mitochondrial translation, DNA repair, and Rho signaling, with CRISPR validation highlighting therapeutic targets. #CancerGenomics #PrecisionMedicine #CancerGenomics #PrecisionMedicine
www.nature.com/articles/s41...

AI-powered liquid biopsies are revolutionizing cancer detection! 🧬✨ Discover how deep learning is unlocking hidden patterns in cell-free DNA (cfDNA) for earlier, more accurate, and non-invasive cancer diagnostics.

Read more 👇

sanjeevareddydodlapati.substack.com/p/revolution...

Am I the only one feeling schadenfreude about USA's arrogance about it's AI superiority, as they got their butt kicked by DeepSeek-R1, and that the facade of needing 100K GPUs for training was an intellectually lazy and uninspiring solution that crumbled before the world's eyes? 👀

2 interns available in my team!! www.linkedin.com/posts/activi...

Ever wondered how the complexity of human biology—from our organs to our very DNA—unfolds at different levels? Let’s decode life together!
#ComputationalGenomics #Genomics #AIinScience #Bioinformatics #HumanBiology #Omics #MolecularBiology #ScienceCommunication
open.substack.com/pub/sanjeeva...

We've a postdoc opening for our lab at the Broad: Cambridge MA!

Must have experience in toxicology + data science

Work on the wonderful OASIS dataset we are producing... Cell Painting, transcriptomics, proteomics in various liver cell and tissue models!

broad.io/mlcbpostdoc

The Alarming Impact of Microplastics on Human Health and the Environment 🧵
1/ Microplastics are everywhere: in our food, water, air, and even inside our bodies. But what are they, where do they come from, and how do they affect us? Let’s dive in. 🧬🌍 #Microplastics #Health #Environment

NEW PREPRINT

A detailed overview of 32 popular predictive performance metrics for prediction models

arxiv.org/abs/2412.10288

Excited to be presenting Orthrus with Ruain Shi and Keren Isaev @karini925.bsky.social today! We will be presenting our spotlight at the workshop on AI for new drug modalities #NeurIPS2024

Come chat about a new approach to mRNA representation learning!

This work finds that costly foundation models are outperformed by CNNs in genomics benchmarks, and are outperformed by *linear* autoregression in time-series benchmarks. 👀

I've been working to make designing regulatory DNA that exhibits desired characteristics easier for everyone. With the following series of tools, you can go from a blank slate to designed edits in ~30 minutes using only a V100. That includes file downloading and model training.

Also want to point to this other recent preprint that also shows that optimized ab-initio CNN models beat DNALMs even on the (relatively pointless) surrogate tasks used in the DNALM papers. CNNs also beat several other foundation models in other domains.

arxiv.org/abs/2411.02796

#bioMLeval

Protein and DNA Conformational Changes Contribute to Specificity of Cre Recombinase https://www.biorxiv.org/content/10.1101/2024.12.11.627928v1

Excited about L2G, led by Wenduo Cheng. We leverage LLMs to beat genomic FMs and strong supervised baselines on a wide range of benchmarks. L2G uses cross-modal transfer (rather than vanilla fine-tuning), and neural architecture search to learn a genomic-specific embedder model.

I trained a sparse autoencoder on the middle layer residual stream of my genome language model and found human-interpretable latent features that consistently detect specific DNA motifs!

🧵1/8