学术报告：Joint triplet sampling enables structure-preserved dimension reduction in single-cell transcriptomic data

报告时间：2023年10月30日（星期一）下午14:00-16:00

报告地点：沙河校区，学院1号楼102会议室

报告人：许欣怡，中央财经大学统计与数学学院，讲师

报告摘要：Dimension reduction (DR) plays an important role in single-cell RNA sequencing analysis. A desired dimension reduction method should be applicable to various application scenarios in scRNA-seq, including identifying cell types, preserving the inherent structure of data and handling with batch effects. However, most existing DR methods fail to accommodate these requirements simultaneously. In this talk, we develop a novel structure-preserved dimension reduction (SPDR) method using intra- and inter-batch triplets sampling, which captures higher-order structure information and meanwhile accounts for batch information of the data. Then we minimize a robust loss function for the chosen triplets to obtain a structure-preserved and batch-corrected low-dimensional representation. Comprehensive evaluations show that SPDR outperforms other competing DR methods in visualization with an authentic gene expression pattern. We believe that SPDR will be a valuable tool for characterizing complex cellular heterogeneity in single-cell transcriptomics.

报告人简介：许欣怡，中央财经大学统计与数学学院讲师。主要研究方向为高维复杂数据分析、单细胞多组学数据分析、深度学习在生物信息中的应用。本科毕业于中央财经大学数学与应用数学系，博士毕业于中国人民大学数理统计学系，曾赴美国宾夕法尼亚大学生物统计学与信息学系访问交流一年。主要研究成果发表于Nature Communications, Briefings in Bioinformatics等生物信息学顶级期刊，以及Journal of Statistical Planning and Inference等统计学期刊。主持全国统计科学研究优选项目等多项课题。