Investigator, Biostatistics Branch
Division of Cancer Epidemiology and Genetics, National Cancer Institute
Statistical challenges in analyzing complex genetic and genomic
Shady Grove TE 406
Dr. Shi will first review his recent work on copy number variation analyses based on SNP arrays, including detecting CNVs in nuclear families, testing associations of recurrent CNVs and CNVs randomly distributed in a short genomic region. Using a recent family-based melanoma study, he will then briefly discuss a chromosome-based exact test for assessing the association of rare genetic variants in family-based studies and an analyses pipeline for inferring the age of founder mutations.
He will then present a comprehensive characterization of the genetic basis of methylome diversity in EAGLE histologically normal human lung tissues. The large number of detected meQTLs are strongly enriched in regulatory regions based on the ENCODE data and enriched for lung cancer risk. In particular, 4 out of the 5 established lung cancer GWAS SNPs are meQTLs but only one is eQTL, suggesting that DNA methylation instead of proximal gene expression is involved in carcinogenesis.
Finally, Dr. Shi will introduce a novel statistical algorithm for detecting master regulators in multi-omics studies for traits (>10,000) with long range dependence. Over 30 master regulatory SNPs in a RNA-seq based eQTL study and one master regulatory SNP in EAGLE meQTL study with replication in TCGA data were detected.