Generative pretraining from large-scale transcriptomes for single-cell deciphering

作     者：Shen, Hongru Liu, Jilei Hu, Jiani Shen, Xilin Zhang, Chao Wu, Dan Feng, Mengyao Yang, Meng Li, Yang Yang, Yichen Wang, Wei Zhang, Qiang Yang, Jilong Chen, Kexin Li, Xiangchun 

作者机构：Tianjin Med Univ Tianjin Med Univ Canc Inst & Hosp Tianjin Canc Inst Tianjins Clin Res Ctr CancNatl Clin Res Ctr Canc Tianjin Peoples R China Tianjin Med Univ Tianjin Med Univ Canc Inst & Hosp Dept Bone & Soft Tissue Tumor Tianjins Clin Res Ctr CancNatl Clin Res Ctr Canc Tianjin Peoples R China Tianjin Med Univ Tianjin Med Univ Canc Inst & Hosp Dept Epidemiol & Biostat Natl Clin Res Ctr CancKey Lab Mol Canc Epidemiol Tianjin Peoples R China Tianjin Med Univ Tianjin Med Univ Canc Inst & Hosp Tianjins Clin Res Ctr Canc Dept Maxillofacial & Otorhinolaryngol OncolNatl C Tianjin Peoples R China 

出 版 物：《ISCIENCE》 (iScience)

年 卷 期：2023年第26卷第5期

页      面：106536页

核心收录：

基　　金：National Key Research and Development Program of China [2021YFC2500400] National Natural Science Foundation of China [32270688, 31801117, 31900471, 82073287] Tianjin Municipal Health Commission Foundation [RC20027] Tianjin Key Medical Discipline (Specialty) Construction Project [TJYXZDXK-009A] 

主　　题：Automation in bioinformatics Data processing in systems biology Transcriptomics 

摘      要：Exponential accumulation of single-cell transcriptomes poses great challenge for efficient assimilation. Here, we present an approach entitled generative pretrain-ing from transcriptomes (tGPT) for learning feature representation of transcrip-tomes. tGPT is conceptually simple in that it autoregressive models the ranking of a gene in the context of its preceding neighbors. We developed tGPT with 22.3 million single-cell transcriptomes and used four single-cell datasets to eval-utate its performance on single-cell analysis tasks. In addition, we examine its ap-plications on bulk tissues. The single-cell clusters and cell lineage trajectories derived from tGPT are highly aligned with known cell labels and states. The feature patterns of tumor bulk tissues learned by tGPT are associated with a wide range of genomic alteration events, prognosis, and treatment outcome of immunotherapy. tGPT represents a new analytical paradigm for integrating and deciphering massive amounts of transcriptome data and it will facilitate the inter-pretation and clinical translation of single-cell transcriptomes.