Deep learning based subdivision approach for large scale macromolecules structure recovery from electron cryo tomograms

TL;DR

This paper introduces a deep learning-based subdivision method for large-scale macromolecular structure recovery from electron cryo tomograms, improving discrimination, scalability, and enabling discovery of new structures.

Contribution

It presents a novel supervised deep learning approach combined with clustering for subdividing subtomograms, enhancing existing methods' ability to handle large, heterogeneous datasets.

Findings

Significant improvement in discrimination ability over previous methods

Enhanced scalability for large datasets

Ability to discover new structural classes

Abstract

Motivation: Cellular Electron CryoTomography (CECT) enables 3D visualization of cellular organization at near-native state and in sub-molecular resolution, making it a powerful tool for analyzing structures of macromolecular complexes and their spatial organizations inside single cells. However, high degree of structural complexity together with practical imaging limitations make the systematic de novo discovery of structures within cells challenging. It would likely require averaging and classifying millions of subtomograms potentially containing hundreds of highly heterogeneous structural classes. Although it is no longer difficult to acquire CECT data containing such amount of subtomograms due to advances in data acquisition automation, existing computational approaches have very limited scalability or discrimination ability, making them incapable of processing such amount of data. Results: To complement existing approaches, in this paper we propose a new approach for subdividing subtomograms into smaller but relatively homogeneous subsets. The structures in these subsets can then be separately recovered using existing computation intensive methods. Our approach is based on supervised structural feature extraction using deep learning, in combination with unsupervised clustering and reference-free classification. Our experiments show that, compared to existing unsupervised rotation invariant feature and pose-normalization based approaches, our new approach achieves significant improvements in both discrimination ability and scalability. More importantly, our new approach is able to discover new structural classes and recover structures that do not exist in training data.