genomics

A new review in the New England Journal of Medicine highlights that polygenic risk scores for various diseases cost only $20 and are informative for high-risk individuals across ancestries, underscoring the underutilized value of genomics in medical practice.

MIST introduces a novel multimodal approach for survival prediction in oncology by using genomic-guided histology attention, showing improved performance across multiple external datasets.

Arc Institute is hiring experts in disease biology, computational biology, functional genomics, and NGS for its Tech Centers and Core Labs, with various roles based in Palo Alto.

Google's @antigravity team has released several updates, including AlphaGenome Atlas Skills for AI-driven genomic discovery, a multi-agent /boost command for developer tasks, and improvements to interactive UI and version control tools.

Samuel King, a Stanford PhD student, is honored as an MIT Technology Review Innovator Under 35 for his work using the AI model Evo to design functional bacteriophages, advancing AI applications in biological design.

DeepMind has released the AlphaGenome Atlas, a public repository of precomputed predictions for all 9 billion possible single-letter DNA variants, providing an interface for scientists to study gene regulation without coding.

Google DeepMind's AlphaGenome Atlas is an AI-powered map of all possible single-letter DNA mutations, providing a 1-petabyte dataset for researchers to explore genetic variants via a free web interface and API.

Google DeepMind introduces AlphaGenome Atlas, a platform with AI predictions for all 9 billion single-letter DNA variants in the human genome, accelerating biological research and disease understanding.

Jingtian Zhou builds single-cell and computational frameworks to probe the 3D genome, contributing to major publications in Science, Cell, and Cancer Cell that advance understanding of cell states and diseases.

This paper introduces Scientific Agent Skills, an open-source library of 163 procedural knowledge skills across 16 scientific domains to help AI research agents perform defensible analyses.

MIT Technology Review profiles Deanne Taylor, a bioinformatics director who pushed for pediatric representation in the Human Cell Atlas and helped launch the dGTEx project to map healthy gene expression in children, addressing the lack of baseline data for pediatric medicine.

Arc Institute announces a Nature publication solving the structural mechanism of bridge excision in bridge RNA-guided DNA rearrangement, providing new engineering handles for programmable biology.

Chinese researchers at BGI-Research released OneGenome, an open-source AI system that interprets gene mutations for clinical diagnosis, outperforming general LLMs like DeepSeek-v4, to help shorten the diagnostic odyssey for rare disease patients.

This paper introduces bioMoR, a biology-guided Mixture-of-Recursions framework for genomic learning that integrates structured biological knowledge into recursive Transformer architectures, improving efficiency and accuracy across omics benchmarks.

A new 1.1B-parameter DNA foundation model, MarinDNA v0.5 scaling ladder, was released on Hugging Face; it reads and generates DNA sequences and reportedly rivals Evo 2 40B on variant effect prediction.

ENCODE GRAMMAR is a deep learning model resource for decoding DNA sequence logic of regulatory elements in the human genome, with a blog series covering model access, interpretation, and applications.

Researchers using ancestral recombination graphs found evidence of a third 'ghost' human lineage in Africa that interbred with modern humans before the out-of-Africa migration.

A new study in Science provides a comprehensive atlas of DNA methylation and 3D genome architecture across human cell types.

GoodfireAI used interpretability to predict which genetic variants cause disease, validated predictions with real-world data, and released an open-source database for all variants in the NIH's ClinVar database.

The author built a proof-of-concept private genomics study using Stoffel MPC, where one hundred simulated participants computed aggregate allele counts without any party seeing the complete dataset, demonstrating privacy-preserving genomic analysis.