{"title":"脚本:预测单细胞远程顺式调节基于预训练的图注意网络。","authors":"Yu Zhang, Baole Wen, Yifeng Jiao, Yuchen Liu, Xin Guo, Yushuai Wu, Jiyang Li, Limei Han, Yinghui Xu, Xin Gao, Yuan Qi, Yuan Cheng, Ying He, Weidong Tian","doi":"10.1002/advs.202505021","DOIUrl":null,"url":null,"abstract":"<p><p>Single-cell cis-regulatory relationships (CRRs) are essential for deciphering transcriptional regulation and understanding the pathogenic mechanisms of disease-associated non-coding variants. Existing computational methods struggle to accurately predict single-cell CRRs due to inadequately integrating causal biological principles and large-scale single-cell data. Here, SCRIPT (Single-cell Cis-regulatory Relationship Identifier based on Pre-Trained graph attention networks) is presented for inferring single-cell CRRs from transcriptomic and chromatin accessibility data. SCRIPT incorporates two key innovations: graph causal attention networks supported by empirical CRR evidence, and representation learning enhanced through pretraining on atlas-scale single-cell chromatin accessibility data. Validation using cell-type-specific chromatin contact and CRISPR perturbation data demonstrates that SCRIPT achieves a mean AUC of 0.89, significantly outperforming state-of-the-art methods (AUC: 0.7). Notably, SCRIPT obtains an over twofold improvement in predicting long-range CRRs (>100 Kb) compared to existing methods. By applying SCRIPT to Alzheimer's disease and schizophrenia, a framework is established for prioritizing disease-causing variants and elucidating their functional effects in a cell-type-specific manner. By uncovering molecular genetic mechanisms undetected by existing computational methods, SCRIPT provides a roadmap for advancing genetic diagnosis and target discovery.</p>","PeriodicalId":117,"journal":{"name":"Advanced Science","volume":" ","pages":"e05021"},"PeriodicalIF":14.1000,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"SCRIPT: Predicting Single-Cell Long-Range Cis-Regulation Based on Pretrained Graph Attention Networks.\",\"authors\":\"Yu Zhang, Baole Wen, Yifeng Jiao, Yuchen Liu, Xin Guo, Yushuai Wu, Jiyang Li, Limei Han, Yinghui Xu, Xin Gao, Yuan Qi, Yuan Cheng, Ying He, Weidong Tian\",\"doi\":\"10.1002/advs.202505021\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>Single-cell cis-regulatory relationships (CRRs) are essential for deciphering transcriptional regulation and understanding the pathogenic mechanisms of disease-associated non-coding variants. Existing computational methods struggle to accurately predict single-cell CRRs due to inadequately integrating causal biological principles and large-scale single-cell data. Here, SCRIPT (Single-cell Cis-regulatory Relationship Identifier based on Pre-Trained graph attention networks) is presented for inferring single-cell CRRs from transcriptomic and chromatin accessibility data. SCRIPT incorporates two key innovations: graph causal attention networks supported by empirical CRR evidence, and representation learning enhanced through pretraining on atlas-scale single-cell chromatin accessibility data. Validation using cell-type-specific chromatin contact and CRISPR perturbation data demonstrates that SCRIPT achieves a mean AUC of 0.89, significantly outperforming state-of-the-art methods (AUC: 0.7). Notably, SCRIPT obtains an over twofold improvement in predicting long-range CRRs (>100 Kb) compared to existing methods. By applying SCRIPT to Alzheimer's disease and schizophrenia, a framework is established for prioritizing disease-causing variants and elucidating their functional effects in a cell-type-specific manner. By uncovering molecular genetic mechanisms undetected by existing computational methods, SCRIPT provides a roadmap for advancing genetic diagnosis and target discovery.</p>\",\"PeriodicalId\":117,\"journal\":{\"name\":\"Advanced Science\",\"volume\":\" \",\"pages\":\"e05021\"},\"PeriodicalIF\":14.1000,\"publicationDate\":\"2025-08-20\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Advanced Science\",\"FirstCategoryId\":\"88\",\"ListUrlMain\":\"https://doi.org/10.1002/advs.202505021\",\"RegionNum\":1,\"RegionCategory\":\"材料科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"CHEMISTRY, MULTIDISCIPLINARY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Advanced Science","FirstCategoryId":"88","ListUrlMain":"https://doi.org/10.1002/advs.202505021","RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"CHEMISTRY, MULTIDISCIPLINARY","Score":null,"Total":0}
SCRIPT: Predicting Single-Cell Long-Range Cis-Regulation Based on Pretrained Graph Attention Networks.
Single-cell cis-regulatory relationships (CRRs) are essential for deciphering transcriptional regulation and understanding the pathogenic mechanisms of disease-associated non-coding variants. Existing computational methods struggle to accurately predict single-cell CRRs due to inadequately integrating causal biological principles and large-scale single-cell data. Here, SCRIPT (Single-cell Cis-regulatory Relationship Identifier based on Pre-Trained graph attention networks) is presented for inferring single-cell CRRs from transcriptomic and chromatin accessibility data. SCRIPT incorporates two key innovations: graph causal attention networks supported by empirical CRR evidence, and representation learning enhanced through pretraining on atlas-scale single-cell chromatin accessibility data. Validation using cell-type-specific chromatin contact and CRISPR perturbation data demonstrates that SCRIPT achieves a mean AUC of 0.89, significantly outperforming state-of-the-art methods (AUC: 0.7). Notably, SCRIPT obtains an over twofold improvement in predicting long-range CRRs (>100 Kb) compared to existing methods. By applying SCRIPT to Alzheimer's disease and schizophrenia, a framework is established for prioritizing disease-causing variants and elucidating their functional effects in a cell-type-specific manner. By uncovering molecular genetic mechanisms undetected by existing computational methods, SCRIPT provides a roadmap for advancing genetic diagnosis and target discovery.
期刊介绍:
Advanced Science is a prestigious open access journal that focuses on interdisciplinary research in materials science, physics, chemistry, medical and life sciences, and engineering. The journal aims to promote cutting-edge research by employing a rigorous and impartial review process. It is committed to presenting research articles with the highest quality production standards, ensuring maximum accessibility of top scientific findings. With its vibrant and innovative publication platform, Advanced Science seeks to revolutionize the dissemination and organization of scientific knowledge.