Protein Graph Neural Networks for Heterogeneous Cryo-EM Reconstruction
This addresses the problem of reconstructing protein structures from cryo-EM data for structural biology, but it is incremental as it builds on existing methods with a novel inductive bias.
The paper tackled heterogeneous cryo-EM reconstruction by developing a geometry-aware method using a graph neural network to predict atomic backbone conformations, achieving higher accuracy than a comparable MLP on synthetic datasets.
We present a geometry-aware method for heterogeneous single-particle cryogenic electron microscopy (cryo-EM) reconstruction that predicts atomic backbone conformations. To incorporate protein-structure priors, we represent the backbone as a graph and use a graph neural network (GNN) autodecoder that maps per-image latent variables to 3D displacements of a template conformation. The objective combines a data-discrepancy term based on a differentiable cryo-EM forward model with geometric regularization, and it supports unknown orientations via ellipsoidal support lifting (ESL) pose estimation. On synthetic datasets derived from molecular dynamics trajectories, the proposed GNN achieves higher accuracy compared to a multilayer perceptron (MLP) of comparable size, highlighting the benefits of a geometry-informed inductive bias.