讲座编号：jz-yjsb-2017-y062
讲座问题：Applications Of Statistics In Genetics/Genomics
主 讲 人：Yudi Pawitan 教授 瑞典斯德哥尔摩卡罗林斯卡研究所
讲座时间：2017年10月15日（周日）上午10：00
讲座所在：阜成路东校区一号楼241 （理学院聚会室）
加入工具：理学院统计专业的研究生及青年西席
主理单位：研究生院
承办单位：理学院
主讲人简介：
Prof. Yudi Pawitan(yudi.pawitan@ki.se) ,professor of department of medical epidemiology and biostatistics,Karolinska Institutet, Stockholm, Sweden.His research interests include statistical genetics,microarrays,family data,biostatistics and likelihood inference.He has published in journals in several fields,including Stat Med,Stat Methods Med Res,Breast Cancer Res,Sci Rep,Genet Epidemiol,Stat Appl Genet Mol Biol,Biostatistics,Brief Bioinform,Oncotarget,Cancer Res,Prostate Cancer Prostatic Dis and Mol Psychiatry.
主讲内容：
Technological advances in high-throughput molecular measurements in genetics and genomics are revolutionizing biological and medical research. The data explosion means that statistics and statisticians have an important role to play because of our expertise in modelling and inference. This talk will be split into three parts: (i) A selection operator for summary association statistics reveals allelic heterogeneity of complex traits, (ii) Pleiotropic insights of 22 novel loci for human anthropometry: discovery, replication, and in silico functional investigation, and (iii) Integration of omics data to discover driver genes in cancer. In part (i), we describe a novel variable-selection procedure to complement the genome-wide association studies (GWAS) in order to estimate the allelic heterogeneity within the significant loci. Part (ii) describes a joint analysis of multiple phenotypes in GWAS. We implement a fast and powerful multi-trait meta-analysis for genetic studies, which is able to combine any set of phenotypes, with any combination of outcome distributions and any level of sample overlap. In part (iii) we show an algorithm to integrate multiple omics data at DNA and RNA levels, together with functional information such as gene-protein networks. A successful integration can reveal potential cancer drivers, which are biologically meaningful biomarkers (i) for better understanding of cancer progression and (i) for clinical translation in terms of therapeutic targets in personalized medicine.