Detection of cell-type-specific risk-CpG sites in epigenome-wide association studies

發布者:文明辦發布時間:2019-06-27瀏覽次數:509


主講人:魏穎穎 香港中文玖体育直播助理教授


時間:2019年6月29日13:00


地點:三號樓332會議廳


舉辦單位:數理學院


主講人介紹:魏穎穎于2009年畢業于清華玖体育直播數學系,2014年在約翰霍普金斯玖体育直播獲得計算機科學碩士學位及生物統計博士學位,同年起在香港中文玖体育直播統計系任助理教授。


內容介紹:In epigenome-wide association studies, the measured signals for each sample are  a mixture of methylation profiles from different cell types. The current  approaches to the association detection only claim whether a  cytosine-phosphate-guanine (CpG) site is associated with the phenotype or not,  but they cannot determine the cell type in which the risk-CpG site is affected  by the phenotype. Here, we propose a solid statistical method, HIgh REsolution  (HIRE), which not only substantially improves the power of association detection  at the aggregated level as compared to the existing methods but also enables the  detection of risk-CpG sites for individual cell types. We prove that the  proposed model is identifiable under conditions easily met in real data.