SeminarsGenomic meta-analysis methods and applications.
68
reads
As many high-throughput genomic technologies continue to thrive, massive data are generated daily and accumulated in the public domain. Development of statistical information integration methods for omics data has become an urgent and challenging task. In this talk, we consider horizontal genomic meta-analysis, which extends from traditional meta-analysis to a genomic scale to combine multiple studies of relevant hypothesis (e.g. microarray studies, GWAS or epigenomic data) and to increase statistical power for more validated conclusions. I will discuss the hypothesis settings behind different genomic meta-analysis methods. I will then present several new methods we have developed recently (including an adaptively weighted Fisher’s method, an order statistic method and a missing value imputation method) and discuss their statistical properties.
reads
Chien-Cheng Tseng
2013-05-22
15:30:00 - 17:00:00
103 , Mathematics Research Center Building (ori. New Math. Bldg.)
As many high-throughput genomic technologies continue to thrive, massive data are generated daily and accumulated in the public domain. Development of statistical information integration methods for omics data has become an urgent and challenging task. In this talk, we consider horizontal genomic meta-analysis, which extends from traditional meta-analysis to a genomic scale to combine multiple studies of relevant hypothesis (e.g. microarray studies, GWAS or epigenomic data) and to increase statistical power for more validated conclusions. I will discuss the hypothesis settings behind different genomic meta-analysis methods. I will then present several new methods we have developed recently (including an adaptively weighted Fisher’s method, an order statistic method and a missing value imputation method) and discuss their statistical properties.