$0of $10,000
This proposal asks whether cells reuse a small set of response programs across genetic and drug perturbations, testing if structured models predict unseen perturbations better and transfer from genes to drugs. The held-out perturbation design is a sound test, but fit to an RFP focused on molecular dynamics and docking is weak, and the required GPU-hour estimate, checkpointing plan, datasets, and model details are all missing.