molecular-simulations

Analyze protein-protein/ligand interactions from MD trajectories with molecular-simulations — per-residue interaction-energy fingerprinting (electrostatic + Lennard-Jones), linear interaction energy (static structure or dynamic trajectory), and KMeans clustering of per-frame feature data. Use when characterizing binding interfaces, footprinting which residues drive an interaction, computing chain-chain interaction energies, or clustering conformations/fingerprints.

Score predicted protein complex interfaces with molecular-simulations' ipSAE — computes ipSAE, ipTM, pDockQ, pDockQ2, and LIS per chain pair from a predicted structure plus its pLDDT/PAE confidence arrays. Use when ranking or filtering AlphaFold-Multimer / Boltz / Chai docking predictions by interface confidence, or batch-scoring many predicted models.

Compute solvent-accessible surface area on a structure or MD trajectory with molecular-simulations — absolute per-residue SASA (Shrake-Rupley) and RelativeSASA fractional exposure (0=buried, 1=exposed). Use when measuring residue burial/exposure, finding solvent-exposed surfaces or buried interfaces, or quantifying conformational changes in accessibility over a trajectory.

Build AMBER molecular systems (topology + coordinates) for OpenMM simulations with molecular-simulations. Use when preparing a protein, protein-ligand complex, or implicit-solvent system from a PDB/CIF — solvating, neutralizing/ionizing to 150 mM NaCl, parameterizing small molecules with GAFF2, or assigning chains/disulfides before running MD.

Configure Parsl to deploy molecular-simulations builds, MD runs, and analyses across HPC. Use when running many systems/replicas in parallel, picking a compute platform (Local single-node GPU, Heterogeneous GPU+CPU, Polaris, or Aurora via PBSPro), pinning workers to GPUs, or writing/loading a YAML settings file for a cluster job.

Run OpenMM molecular dynamics with molecular-simulations — explicit-solvent NPT, implicit-solvent GB, energy minimization, and MM-PBSA binding free energy. Use when launching, configuring, or restarting an MD production run from AMBER/CHARMM inputs, choosing equilibration/production step counts, or selecting GPU/CPU platforms. For running many replicas across HPC nodes, combine with the parsl-hpc skill.