1.李扬 教授 中国人民大学 9:00-10:00
题目：Pairwise Sequential Randomization and Its Properties
摘要：This article introduces a new randomization procedure to improve the covariate balance across treatment groups. Covariate balance is one of the most important concerns for successful comparative studies, such as causal inference and clinical trials, because it reduces bias and improves the accuracy of inference. However, chance imbalance may still exist in traditional randomized experiments, in which units are randomly allocated without using their covariate information. To address this issue, the proposed method allocates the units sequentially and adaptively, using information on the current level of imbalance and the incoming unit's covariate. With a large number of covariates or a large number of units, the proposed method shows substantial advantages over the traditional methods in terms of the covariate balance and computational time, making it an ideal technique in the era of big data. Furthermore, the proposed method attains the optimal covariate balance, in the sense that the estimated average treatment effect under the proposed method attains its minimum variance asymptotically. Numerical studies and real data analysis provide further evidence of the advantages of the proposed method.
2.吴梦云 副教授 上海财经大学 10:00-11:00
题目：Integrative interaction analysis of multi-omics data
摘要：For the etiology, progression, and treatment of complex diseases, gene-environment (G-E) interactions have important implications beyond the main G and E effects. With the small sample sizes of omics profiling studies and noisy nature of omics data, the interaction analysis on a single dataset often leads to unsatisfactory results. In recent profiling studies, a prominent trend is to collect measurements on multi-omics data, including gene expressions as well as their regulators (copy number alteration, microRNA, methylation, etc.) on the same subjects. In our study, we propose a joint interaction analysis approach in an integrative perspective based on the biclustering and regularized estimation techniques, uniquely effectively accommodating the regulation relationships among multi-omics data. Simulations show that the proposed approach has significantly improved identification performance. In the analysis of cancer multi-omics data, biologically sensible findings different from the alternatives are made.
3.史兴杰 副教授 南京财经大学 11:00-12:00
题目：Variational Inference for Multiple Correlated Outcomes in Large Scale Data
For large-scale inference, where multiple correlated outcomes have been measured on samples, a joint analysis strategy, whereby the outcomes are analyzed jointly, can improve statistical power over a single-outcome analysis strategy. There are two questions of interest to be addressed when conducting variable selection with multiple traits. The first question examines whether a feature is significantly associated with any of the outcomes being tested. The second question focuses on identifying the specific variable(s) that is associated with the outcome. Since existing methods primarily focus on the first question, this paper seeks to provide a complementary method that addresses the second question.
In this talk, I will discuss about a novel method, Variational Inference for Multiple Correlated Outcomes (VIMCO), that focuses on identifying the specific response that is associated with the variable, when performing a joint analysis of multiple outcomes, while accounting for correlation among the multiple outcomes. We performed extensive numerical studies and also applied VIMCO to analyze two GWAS datasets. The numerical studies and real data analysis demonstrate that VIMCO improves statistical power over single-trait analysis strategies when the multiple traits are correlated and has comparable performance when the traits are not correlated.