Journal of computational biology: A journal of computational molecular cell biology | |
Attentive Cross-Modal Paratope Prediction | |
AndreeaDeac^1,21  PetarVeliČković^23  PietroSormanni^3,44  | |
[1] Address correspondence to: Andreea Deac, BA, Department of Computer Science and Technology, University of Cambridge, 15 JJ Thomson Avenue, Cambridge CB3 0FD, United Kingdom^1;Department of Chemistry, University of Cambridge, Cambridge, United Kingdom^4;Department of Computer Science and Technology and University of Cambridge, Cambridge, United Kingdom^2;Dr. Pietro Sormanni, Centre for Misfolding Diseases , Department of Chemistry, University of Cambridge, Cambridge CB2 1EW, United Kingdom^3 | |
关键词: àtrous; antibody; antigen; attention; cross-modal; paratope; | |
DOI : 10.1089/cmb.2018.0175 | |
学科分类:生物科学(综合) | |
来源: Mary Ann Liebert, Inc. Publishers | |
【 摘 要 】
Antibodies are a critical part of the immune system, having the function of recognizing and mediating the neutralization of undesirable molecules (antigens) for future destruction. Being able to predict which amino acids belong to theparatope, the region on the antibody that binds to the antigen, can facilitate antibody engineering and predictions of antibody-antigen structures. The suitability of deep neural networks has recently been confirmed for this task, with Parapred outperforming all prior models. In this work, we first significantly outperform the computational efficiency of Parapred by leveraging à trous convolutions and self-attention. Second, we implementcross-modal attentionby allowing the antibody residues to attend over antigen residues. This leads to new state-of-the-art results in paratope prediction, along with novel opportunities to interpret the outcome of the prediction.
【 授权许可】
Unknown
【 预 览 】
Files | Size | Format | View |
---|---|---|---|
RO201910256289947ZK.pdf | 3518KB | download |