Live Intelligence
What the biotech AI community is talking about right now
Prof. of Biology, MIT
Boltz-2 now handles covalent ligands and glycosylated proteins out of the box. Ran it on a notoriously difficult kinase–inhibitor complex — RMSD under 1.2 Å. The pace of open-source structural biology is unreal.
Nobel Laureate, UW IPD
New RFdiffusion checkpoint trained on 150K experimental structures + 2M AF2 predictions. Binder design success rates jumped from ~15% to ~40% in our wet-lab validation. Preprint dropping Friday.
Director, Scripps Translational Science
Recursion's Phase II results for their AI-discovered STING agonist just dropped. ORR of 38% in refractory solid tumors. This is the first AI-native drug to show real clinical signal in a randomized trial. Landmark moment.
CEO, insitro
We're releasing our multi-modal foundation model for cell biology — trained on 40M cell images paired with transcriptomics. It predicts phenotypic outcomes from perturbations with 0.87 AUROC. Open weights on HuggingFace.
Founder, DeepLearning.AI
Just visited Insilico Medicine's Qingdao lab. Their AI-designed anti-fibrotic (INS018_055) is in Phase III now. What struck me: the entire lead optimization took 8 months vs the typical 3-4 years. AI is compressing timelines dramatically.
Prof. of Systems Biology, Columbia
Hot take: protein language models have plateaued on standard benchmarks. The next frontier is dynamics — predicting conformational ensembles, not single structures. Whoever cracks this wins the next decade of drug discovery.
Prof. of Biomedical Informatics, Harvard
Our new graph foundation model for drug repurposing just identified 3 existing approved drugs with predicted activity against treatment-resistant AML. Two validated in patient-derived xenograft models. Paper in Nature Medicine next week.
CSO, AI Proteins
RFantibody v2 is live. Redesigned the CDR-H3 sampling — humanization scores are now consistently >90% while maintaining binding affinity. Also added nanobody support. Check the repo.
GP, a16z Bio + Health
Just led a $200M Series C into Xaira Therapeutics. Their generative chemistry platform has 4 programs in IND-enabling studies. The thesis: AI doesn't just find drugs faster, it finds drugs that couldn't be found otherwise.
Prof. of Biology, MIT
ColabFold just crossed 2 million runs. When we launched it, I hoped maybe a few thousand people would use it. The democratization of structural biology is the thing I'm most proud of in my career so far.
CEO, Verge Genomics
Our ALS program, VRG50635, just dosed the first patient in Phase I. Entirely identified through our CONVERGE AI platform — from target discovery (PIKfyve) to candidate selection in 18 months. The future is here.
CSO, Relay Therapeutics
We benchmarked ADMET-AI against our internal models on 12 endpoints. It wins on 8 of them. The era of every pharma company building proprietary ADMET models from scratch is ending. Open-source wins again.
Nobel Laureate, UW IPD
Protein degrader design is the next frontier for computational biology. We're combining RFdiffusion with molecular glue modeling to design PROTACs computationally. Early results are very promising — paper in review.
CEO, insitro
Excited to announce: insitro and Eli Lilly expanding our NASH collaboration. Our ML models identified a novel target combination that showed 60% reduction in liver fibrosis in preclinical models. This is precision medicine at scale.