Author and run BioPipelines (locbp-uzh, CSBJ 2026) computational protein & ligand design pipelines on a GPU. BioPipelines gives ~76 comp-bio tools one common Pipeline/Resources API: structure prediction and docking (AlphaFold, Boltz2, ESMFold2, DiffDock, GNINA, NeuralPLexer, DynamicBind), de novo generation (RFdiffusion 1/2/3, RFdiffusionAllAtom, BoltzGen, PocketGen, HBDesigner), inverse folding / sequence design (ProteinMPNN, LigandMPNN, LASErMPNN, Frame2Seq, ThermoMPNN), MD (OpenMM), pocket detection (FPocket, P2Rank, AF2BIND), cheminformatics (RDKit, OpenBabel, ADMET-AI), and interaction/stability analysis (PLIP, ProLIF, PoseBusters, Prodigy). Use for protein binder/enzyme design, inverse folding, ligand docking, compound-library (incl. covalent) screening, and codon optimization. Load when the user wants a BioPipelines workflow, names one of these tools, or wants to run such a design/screening pipeline on GPU.
2026-07-16