FoldHSphere: deep hyperspherical embeddings for protein fold recognition

Current state-of-the-art deep learning approaches for protein fold recognition learn protein embeddings that improve prediction performance at the fold level. However, there still exists aperformance gap at the fold level and the (relatively easier) family level, suggesting that it might be possible...

Full description

Saved in:
Bibliographic Details
Published in:BMC bioinformatics 2021-10, Vol.22 (1), p.1-490, Article 490
Main Authors: Villegas-Morcillo, Amelia, Sanchez, Victoria, Gomez, Angel M
Format: Article
Language:English
Subjects:
Amino acid sequence
Amino acids
Analysis
Classification
Computer architecture
Datasets
Deep learning
Deep neural networks
Embedding
Embedding learning
Hyperspherical space
Machine learning
Methods
Network management systems
Neural networks
Protein fold recognition
Protein folding
Protein structure prediction
Proteins
Prototypes
Recognition
Residual convolutions
Thomson problem
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Current state-of-the-art deep learning approaches for protein fold recognition learn protein embeddings that improve prediction performance at the fold level. However, there still exists aperformance gap at the fold level and the (relatively easier) family level, suggesting that it might be possible to learn an embedding space that better represents the protein folds. In this paper, we propose the FoldHSphere method to learn a better fold embedding space through a two-stage training procedure. We first obtain prototype vectors for each fold class that are maximally separated in hyperspherical space. We then train a neural network by minimizing the angular large margin cosine loss to learn protein embeddings clustered around the corresponding hyperspherical fold prototypes. Our network architectures, ResCNN-GRU and ResCNN-BGRU, process the input protein sequences by applying several residual-convolutional blocks followed by a gated recurrent unit-based recurrent layer. Evaluation results on the LINDAHL dataset indicate that the use of our hyperspherical embeddings effectively bridges the performance gap at the family and fold levels. Furthermore, our FoldHSpherePro ensemble method yields an accuracy of 81.3% at the fold level, outperforming all the state-of-the-art methods. Our methodology is efficient in learning discriminative and fold-representative embeddings for the protein domains. The proposed hyperspherical embeddings are effective at identifying the protein fold class by pairwise comparison, even when amino acid sequence similarities are low.
ISSN:1471-2105
1471-2105
DOI:10.1186/s12859-021-04419-7