" Genome-wide Association Study of Neuroblastoma and High Dimensional Statistics"
Hongzhe Li, Ph.D. - University of Pennsylvania
September 21, 2010 @ 4:00 - 4:30 pm
Location: BRB II/III - Room 253
Biostatistics
TITLE: Genome-wide Association Study of Neuroblastoma and High Dimensional Statistics
ABSTRACT: Neuroblastoma
(NB) is a pediatric cancer of the developing sympathetic nervous system that
most commonly affects young children and is often lethal. It is regarded as a
complex disease with multiple genes contributing to its susceptibility. Recent
genome-wide association study (GWAS) of NB at the Children's Hospital of
Philadelphia (CHOP) has successfully identified several important novel genetic
variants (SNPs) and copy number variants (CNVs) that are associated with NB. We
will first review results from analysis of the CHOP GWAS data, pointing out
some statistical issues that we have to deal with. Motivated by the GWAS of NB, we will then
present several interesting problems in high dimensional statistics that
require further methodological developments, including theoretical investigation
of the GWAS power and sparse signal discovery. We establish the theoretical
boundary for almost recovery of both disease-associated SNPs and non disease-associated
SNPs. We also present optimal segment identification procedures for CNV
analysis based on both SNP-chip data and next generation sequence data. Results
from analysis of NB GWAS are used to illustrate these ideas.
