Skip to main content

ORIGINAL RESEARCH article

Front. Genet.
Sec. Computational Genomics
Volume 15 - 2024 | doi: 10.3389/fgene.2024.1346784
This article is part of the Research Topic Harnessing TCR - Peptide/MHC binding Specificity to Tackle Disease View all 9 articles

TCRcost: A Deep Learning Model Utilizing TCR 3D Structure for Enhanced TCR-peptide Binding Prediction

Provisionally accepted
Fan Li Fan Li Xinyang Qian Xinyang Qian Yuqian Liu Yuqian Liu Xin Lai Xin Lai Xuanping Zhang Xuanping Zhang Xiaoyan Zhu Xiaoyan Zhu *Jiayin Wang Jiayin Wang *
  • School of Computer Science and Technology, Xi’an Jiaotong University, Xi'an, China

The final, formatted version of the article will be published soon.

    Predicting TCR-peptide binding is a complex and significant computational problem in systems immunology. During the past decade, a series of computational methods have been developed for better predicting TCR-peptide binding from amino acid sequences. However, the performance of sequence-based methods appears to have hit a bottleneck. Considering the 3D structures of TCRpeptide complexes, which provide much more information, could potentially lead to better prediction outcomes. In this study, we developed TCRcost, a deep learning methods, to predict TCR-peptide binding by incorporating 3D structures. TCRcost overcomes two significant challenges: acquiring a sufficient number of high-quality TCR-peptide structures and effectively extracting information from these structures for binding prediction. TCRcost corrects TCR 3D structures generated by protein structure tools, significantly extending the available datasets. The main and side chains of a TCR structure are separately corrected using a long short-term memory (LSTM) model. This approach prevents interference between the chains and accurately extracts interactions among both adjacent and global atoms. A 3D convolutional neural network (CNN) is designed to extract the atomic features relevant to TCR-peptide binding. The spatial features extracted by the 3D CNN are then processed through a fully connected layer to estimate the probability of TCR-peptide binding. Testing results demonstrated that predicting TCR-peptide binding from TCR 3D structures is both efficient and highly accurate with an average accuracy of 0.974 on precise structures. Furthermore, the average accuracy on corrected structures was 0.762, significantly higher than the 0.375 accuracy on uncorrected original structures. Additionally, the average root-mean-square-distance (RMSD) to precise structures was significantly reduced from 12.753 Å with predicted structures to 8.785 Å with corrected structures. Thus, utilizing structural information of TCR-peptide complexes is a promising approach for improving binding prediction accuracy.

    Keywords: Systems Immunology, T cell receptor, Peptide binding, Prediction model, protein 3D structure, deep learning, 3D convolutional neural network

    Received: 29 Nov 2023; Accepted: 05 Sep 2024.

    Copyright: © 2024 Li, Qian, Liu, Lai, Zhang, Zhu and Wang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.

    * Correspondence:
    Xiaoyan Zhu, School of Computer Science and Technology, Xi’an Jiaotong University, Xi'an, China
    Jiayin Wang, School of Computer Science and Technology, Xi’an Jiaotong University, Xi'an, China

    Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.