Problem Statement

Current polymer forcefields are generally limited in use to commonly encountered polymers. While very accurate, these forcefields are generally inflexible and cannot be used on novel polymers without retraining. In contrast, the OpenForceField Toolkit allows for the parameterization and simulation of almost all small molecules that may construct a polymer. Using the existing small molecule forcefields to estimate the parameters and charges of repeating subunits in a larger polymer would allow for faster simulation of novel systems.

Scope

This Project only focuses on the very start of adapting the OpenForceField Toolkit to simulate polymers, first developing and testing software infrastructure that can later be used in a scientific study. Namely, we will develop a workflow to load and/or create arbitrary polymers from a minimum amount of information about monomer chemical info and residue connectivity. This involves recognizing the many combinations of ways monomers may arrange themselves and encoding these arrangements in “substructures” that contain chemical information and context to neighboring monomers. This allows functionality like loading a polymer with chemical information from a PDB, building an atomistic polymer from a residue graph, randomly generating polymers, etc.

For now, any discussion of parameter assignment or charge generate is out of scope.

Milestones/Objectives

Links