Foundation Models

2 posts

Ackee-branded molecule viewer showing a protein-DNA complex rendered from an OpenFold3 prediction returned by NVIDIA AI Enterprise, with a voice assistant panel and per-residue walkthrough controls

AI Drug Discovery·April 19, 2026·6 min read

OpenFold3 on NVIDIA AI Enterprise: A Voice-Guided Molecule Viewer in Your Browser

OpenFold3 predicts the 3D structure of any protein complex. NVIDIA AI Enterprise makes it callable as an API. I wired them together into a voice-guided molecule viewer — so anyone, not just structural biologists, can explore a folded protein.

Stack diagram showing the four layers AI drug discovery depends on: a data base layer of genomic, clinical, and experimental data; a pathway layer with cell-type-specific and time-resolved edges; a digital twins and simulation layer with virtual cells and organ-scale models; and AI drug discovery outcomes including target discovery, lead optimization, and in-silico Day Zero success. On the right, a multi-omic foundation model functions as an AlphaFold of pathway state.

AI Drug Discovery·April 18, 2026·4 min read

The Pathway Layer: What AI Drug Discovery Needs to Build Next

AI drug discovery's biggest gap isn't models — it's pathway infrastructure. Six investments biology needs before AI can deliver on its promise.