ProtoMol is an object-oriented, component based, framework for molecular dynamics (MD) simulations. The framework supports the CHARMM 19 and 28a2 force fields and is able to process PDB, PSF, XYZ and DCD trajectory files. It is designed for high flexibility, easy extendibility and maintenance, and high performance demands, including parallelization. The technique of multiple time-stepping is used to improve long-term efficiency. The use of fast electrostatic force evaluation algorithms like Ewald, particle Mesh Ewald (PME), and Multigrid (MG) summation further enhances performance. Longer time steps are possible using MOLLY, Langevin Molly and Hybrid Monte Carlo, Nose-Hoover, and Langevin integrators.

Key Features of ProtoMol 3.0 (available Summer 2009):

1) Interface to OpenMM, an MD library with NVIDIA and ATI general purpose GPU support. OpenMM supports AMBER force fields and Generalized-Born implicit solvent.

2)Python bindings offered as MDLab, which allow for prototyping of high level sampling protocols, new integrators, and new force calculations in Python.

3) Coarse grained normal mode analysis (CNMA), which provides a scalable O(N9/5) time and O(N3/2) memory diagonalization. CNMA approximates low frequency modes very well.

4) Normal Mode Langevin (NML) dynamics, which uses CNMA to periodically compute low frequency bases for propagation of dynamics, while fast modes are minimized to their equilibrium position. NML allows timesteps of 100 fs and more for even small proteins (> 30 residues) with real speedups that are about a third of the timestep used.

5) Full checkpointing support, which simplifies use in distributed computing platforms such as Condor or Folding@Home.