Graph Neural Network Based Coarse-Grained Mapping Prediction

TL;DR

This paper introduces DSGPM, a graph neural network model that automates the prediction of coarse-grained mapping operators for molecular dynamics, outperforming existing methods and producing effective simulation models.

Contribution

The paper presents a novel GNN-based predictor trained on expert-annotated data, using a new metric learning approach for improved graph segmentation in CG mapping.

Findings

DSGPM outperforms state-of-the-art graph segmentation methods.

Predicted CG mappings lead to effective molecular dynamics simulations.

The HAM dataset facilitates future research in CG mapping prediction.

Abstract

The selection of coarse-grained (CG) mapping operators is a critical step for CG molecular dynamics (MD) simulation. It is still an open question about what is optimal for this choice and there is a need for theory. The current state-of-the art method is mapping operators manually selected by experts. In this work, we demonstrate an automated approach by viewing this problem as supervised learning where we seek to reproduce the mapping operators produced by experts. We present a graph neural network based CG mapping predictor called DEEP SUPERVISED GRAPH PARTITIONING MODEL(DSGPM) that treats mapping operators as a graph segmentation problem. DSGPM is trained on a novel dataset, Human-annotated Mappings (HAM), consisting of 1,206 molecules with expert annotated mapping operators. HAM can be used to facilitate further research in this area. Our model uses a novel metric learning objective to produce high-quality atomic features that are used in spectral clustering. The results show that the DSGPM outperforms state-of-the-art methods in the field of graph segmentation. Finally, we find that predicted CG mapping operators indeed result in good CG MD models when used in simulation.