Coarse-grained (CG) protein models have long been studied; indeed, the earliest modeling of protein folding and assembly largely started with CG approaches due to the prohibitive cost of all-atom simulations for the timescales of such processes.  While many of these early efforts intuited CG models and in effect produced “toy” models for capturing general physical behaviors, recently many groups have explored the use of bottom-up coarse-graining methods to reach CG protein models with significantly improved accuracy.  In this talk, we discuss recent advances in using relative entropy coarse-graining to generate new types of CG protein models with quantitative reproduction of thermodynamic and structural properties.  Here, the relative entropy measures the information lost upon coarse graining and is a natural thermodynamic measure of the success of a CG procedure; its minimization provides a universal variational principle for coarse-graining.  We describe the adaptation of relative entropy minimization to create CG peptide models that capture a variety of self-assembly behaviors.  We also describe progress towards CG protein models capable of structurally accurate folding that leverage a new extended ensemble variation of the relative entropy approach.<br/> <br/>

Relative entropy design of coarse-grained protein models

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.