报告主题：The quantile analysis of gene expression trajectory based on multivariate

asymmetric Laplace model

报 告 人：Dianliang Deng 教授

报告时间：2024年11月27日（周三）上午10：00-11：00

报告地点：腾讯会议（会议号：640191788）

报告摘要: Temporal gene expression data provide valuable insights into gene functions, and understanding expression trajectories under various biological conditions has garnered significant scientific interest. This talk introduces a likelihood-based EM algorithm designed to estimate the marginal conditional quantiles of multivariate responses within a linear regression framework. By focusing on associations between multivariate responses and explanatory variables across different quantiles, this approach provides a more robust alternative to traditional methods. We derive the maximum likelihood estimates (MLEs) for the parameters and implement the EM algorithm to perform the estimation. Simulation studies validate the proposed method, demonstrating its effectiveness in parameter estimation. Furthermore, we apply this approach to a real dataset, analyzing the expression of 18 genes in Pseudomonas aeruginosa under 24 biological conditions.

报告人简介：Dr. Deng is a full professor in the Department of Mathematics and Statistics at the University of Regina. His research primarily focuses on the theoretical aspects of probability and statistics, with applications in the biological and medical sciences. He has authored over 50 peer-reviewed academic papers in prestigious journals, including Journal of the Royal Statistical Society (B), Statistica Sinica, etc. Additionally, He serves as a reviewer and referee for more than 20 academic journals, such as Journal of Nonparametric Statistics, Biometrics, etc.