文献类型：期刊文献

英文题名：Sliding-attention transformer neural architecture for predicting T cell receptor-antigen-human leucocyte antigen binding

作者：Feng, Ziyan[1];Chen, Jingyang[2];Hai, Youlong[3];Pang, Xuelian[1];Zheng, Kun[3];Xie, Chenglong[4];Zhang, Xiujuan[4];Li, Shengqing[4];Zhang, Chengjuan[5];Liu, Kangdong[6];Zhu, Lili[1];Hu, Xiaoyong[3];Li, Shiliang[7];Zhang, Jie[2];Zhang, Kai[8];Li, Honglin[1,7,9]

机构：[1]East China Univ Sci & Technol, Sch Pharm, Shanghai Key Lab New Drug Design, Shanghai, Peoples R China;[2]Fudan Univ, Inst Sci & Technol Brain Inspired Intelligence, Shanghai, Peoples R China;[3]Shanghai Jiao Tong Univ, Dept Urol, Shanghai Sixth Peoples Hosp, Sch Med, Shanghai, Peoples R China;[4]Fudan Univ, Huashan Hosp, Dept Pulm & Crit Care Med, Shanghai, Peoples R China;[5]Zhengzhou Univ, Henan Canc Hosp, Ctr Biorepository, Affiliated Canc Hosp, Zhengzhou, Peoples R China;[6]Zhengzhou Univ, Sch Basic Med Sci, Dept Pathophysiol, Zhengzhou, Peoples R China;[7]East China Normal Univ, Innovat Ctr Artificial Intelligence & Drug Discove, Shanghai, Peoples R China;[8]East China Normal Univ, Sch Comp Sci & Technol, Shanghai, Peoples R China;[9]Lingang Lab, Shanghai, Peoples R China

年份：2024

卷号：6

期号：10

起止页码：1216

外文期刊名：NATURE MACHINE INTELLIGENCE

收录：;EI(收录号：20244017121454);WOS:【SCI-EXPANDED(收录号:WOS:001321579200001)】；

基金：We thank Shanghai Applied Protein Technology for supporting the next-generation sequencing processing. This work was supported in part by the National Key Research and Development Program of China (2022YFC3400501); the National Natural Science Foundation of China (82425104, 81825020 and 82150208 to H.L.; 62276099 to K.Z.; 82173690 to Shiliang Li); the Shanghai Rising-Star Program (23QA1402800 to Shiliang Li); the National Program for Special Supports of Eminent Professionals (to H.L.); and the National Program for Support of Top-notch Young Professionals (to H.L.).

语种：英文

外文关键词：Antigens - Cytology

摘要：Neoantigens are promising targets for immunotherapy by eliciting immune response and removing cancer cells with high specificity, low toxicity and ease of personalization. However, identifying effective neoantigens remains difficult because of the complex interactions among T cell receptors, antigens and human leucocyte antigen sequences. In this study, we integrate important physical and biological priors with the Transformer model and propose the physics-inspired sliding transformer (PISTE). In PISTE, the conventional, data-driven attention mechanism is replaced with physics-driven dynamics that steers the positioning of amino acid residues along the gradient field of their interactions. This allows navigating the intricate landscape of biosequence interactions intelligently, leading to improved accuracy in T cell receptor-antigen-human leucocyte antigen binding prediction and robust generalization to rare sequences. Furthermore, PISTE effectively recovers residue-level contact relationships even in the absence of three-dimensional structure training data. We applied PISTE in a multitude of immunogenic tumour types to pinpoint neoantigens and discern neoantigen-reactive T cells. In a prospective study of prostate cancer, 75% of the patients elicited immune responses through PISTE-predicted neoantigens. Predicting TCR-antigen-human leucocyte antigen binding opens the door to neoantigen identification. In this study, a physics-inspired sliding transformer (PISTE) system is used to guide the positioning of amino acid residues along the gradient field of their interactions, boosting binding prediction accuracy.