加入收藏 | 后台管理 | 学校主页

比利时根特大学潘小勇助理研究员来校讲学预告

作者: 时间:2019-08-20 点击数:

应我校计算机与信息安全学院邀请、广西可信软件重点实验室、广西图像图形智能处理重点实验室邀请,比利时根特大学大学潘小勇助理研究员将于822日来我校交流,欢迎广大师生踊跃参与。

讲座题目:Inferring disease-associated microRNAs using semi-supervised multi-label graph convolutional networks

讲座时间:2019822日(周四)上午1000-1130

讲座地点:金鸡岭校区7217

报告摘要:MicroRNAs (miRNAs) play crucial roles in many biological processes involved in diseases. The associations between diseases and protein coding genes (PCGs) have been well investigated, and further the miRNAs in-teract with PCGs to trigger them to be functional. Thus, it is imperative to computationally infer disease-miRNA associations under the context of interaction networks. In this study, we present a computational method, DimiG, to infer miRNA-associated diseases using semi-supervised Graph Convolutional Network model (GCN). DimiG is a multi-label framework to integrate PCG-PCG interactions, PCG-miRNA interactions, PCG-disease associations and tissue expression profiles. DimiG is trained on disease-PCG associations and a graph constructed from interaction networks of PCG-PCG and miRNA-PCG using semi-supervised GCN, which is further used to score associations between diseases and miRNAs. We evaluate DimiG on a benchmark set collected from verified disease-miRNA associations. Our results demonstrate that the new DimiG yields promising performance and outperforms the best published baseline method not trained on disease-miRNA associations by 11% and is also comparable to two state-of-the-art supervised methods trained on disease-miRNA associations. Three case studies of prostate cancer, lung cancer and Inflammatory bowel disease further demonstrate the efficacy of DimiG, where the top miR-NAs predicted by DimiG for them are supported by literature or databases.

报告人简介:潘小勇,比利时根特大学助理研究员,荷兰鹿特丹大学医疗信息学部博士后,硕士毕业于上海交通大学、博士毕业于丹麦哥本哈根大学。研究方向主要包括:电子病历数据分析、生物信息学、机器学习、文本挖掘,在应用机器学习解决生物医学等领域的问题有丰富的科研经验。BioinformaticsComputational biology and Chemistry, Journal of Biomedical and Health Informatics, Interdisciplinary Sciences: Computational Life Sciences等期刊审稿人,已在国际知名期刊发表SCI检索论文30余篇。

2014-2019 © 广西图像图形与智能处理重点实验室
通讯地址:广西壮族自治区桂林市七星区金鸡路1号
邮编:541004