The front page of AI-driven biology
Thursday, October 15, 2026Thu, Oct 15
Updated 7:01 PM PT
Researchers built NoiTox, a computational framework that pairs standard tree-based classifiers using physicochemical properties with deep protein language models and graph networks to predict peptide toxicity.
Researchers from the University of Toronto, the Sloan Kettering Institute, King's College London, and the Francis Crick Institute applied a computational tool named MicroExonator to over 100 terabytes of RNA sequencing data, uncovering more than 1,500 tiny protein-coding segments known as microexons.
Illumina announced the launch of BioInsight, a unified software suite that combines its multiomics analytics, artificial intelligence tools, and biological datasets.
Researchers built RARE-PHENIX, an automated framework that reads unstructured clinical text, standardizes symptoms into formal Human Phenotype Ontology terms, and ranks them by diagnostic importance.
AstraZeneca is paying insitro $40 million upfront, with up to $1.46 billion in potential milestones that bring the deal's headline value to $1.5 billion, to identify drug targets for metabolic liver disease and heart failure.
Conditional generative models frequently hallucinate non-additive interactions when predicting how interventions behave in combination, according to an evaluation across drug dose-response, cellular perturbation, and molecular shape datasets.
Iambic Therapeutics priced its initial public offering at $16 per share to raise $150 million, selling roughly 9.4 million shares ahead of trading on the Nasdaq under ticker IAM.
Researchers used AI-predicted structural models of an antibody-antigen complex to design point mutations that increased the binding affinity of a synthetic single-domain antibody eightfold against interleukin-12.
A computational drug-repurposing pipeline pairs multi-tissue RNA sequencing from treated rodents with a harmonized library of over 900 human disease expression profiles.
Computational biologists built a random forest classifier to predict cancer resistance mechanisms against synthetic lethal therapies using protein interaction networks, gene expression, and cellular essentiality profiles.
Researchers have released MassNet, an AI-ready training dataset containing 1.5 billion tandem mass spectrometry spectra across 35 species, stored in a tensor format designed for deep learning.
Investigators developed Borzoi-SE, a genomic transformer framework that detects super-enhancers, large regulatory DNA clusters that drive cell identity, directly from long-range sequence across 90 human lineages.
A reinforcement learning framework called Multi-turn Feedback Guided optimization trains large language models to iteratively redesign molecules based on feedback from prior modification steps.
Researchers introduced STAMP, a molecular representation format that explicitly separates chemical fragments, their attachment sites, and the overall molecular skeleton.
CloneOT uses lineage tracing to learn custom cost functions for neural optimal transport, predicting how individual cancer clones change after therapeutic perturbation.
Gilead Sciences committed $90 million upfront, comprising $65 million in cash and a $25 million equity investment at a 30% premium, to license three masked T-cell engager therapies engineered using Vir Biotechnology's PRO-XTEN platform.
Researchers combined fragment-based generative molecular design with alchemical absolute binding free-energy calculations to design, prioritize, and synthesize novel small-molecule inhibitors targeting LasB, a virulence enzyme from Pseudomonas aeruginosa.
Zealand Pharma and Flagship Pioneering spinout Generate Biomedicines agreed to collaborate on developing treatments for obesity using generative AI, according to Bloomberg News.
Researchers built two neural network tools, xNNDriver and xAEDriver, to prioritize cancer driver genes by connecting mutations to gene essentiality profiles from the Cancer Dependency Map.
Researchers developed EcoNet, a deep-learning model using graph attention networks that predicts patient response to immunotherapy by modeling cell-cell signaling circuits inside the tumor microenvironment.
Researchers benchmarked generative design pipelines to create non-oxime reactivators of human acetylcholinesterase, an essential enzyme disabled by organophosphate nerve agents and pesticides.
Large language models used to suggest new ionizable lipid molecules showed no meaningful edge over standard genetic algorithms or random search under matching experimental budgets.
Researchers combined interpretable machine learning with modular chemistry to screen degradable ionizable lipids, picking out a lead candidate that successfully delivered mRNA to the liver in mice.
Google DeepMind and Adaptyv Bio published a study in Nature demonstrating SynthIDBio, a watermarking method that embeds identifiable origin tags into AI-generated protein sequences and structures.
Computational binder screening frequently filters candidates by predicted interface confidence, but a benchmark of 196 nanobody loop designs targeting the cancer protein survivin found no correlation between confidence metrics and actual epitope contact (rho = +0.100).
Researchers at Tsinghua University built PeTFlow, a generative flow-matching model that designs binding peptides directly within three-dimensional protein pockets.
A new evaluation framework named Cause-excluded Effect Evaluation exposes that previous single-cell target discovery benchmarks were inflated by data leakage and by leaking the target gene's own expression.
Researchers have introduced SpliceAI2, a deep learning model trained across multiple mammalian species that reconstructs full-length transcripts directly from DNA sequences.
Researchers developed DEPICT, a deep learning model trained on 231,880 cell line-drug pairs across 941 cancer cell lines and 289 compounds to rank which therapies work best inside a specific patient tumor profile.
Researchers compared the specialized splicing model OpenSpliceAI against the genomic foundation model AlphaGenome to predict rare, disease-linked RNA splicing errors called cryptic exons caused by TDP-43 dysfunction.