Application of Pathway-PDT to Identify the Underlying Genetics of Autism Spectrum Disorder

• Activity Number: 190
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): YoSon Park*+ and Eden R. Martin and Michael A. Schmidt and Margaret A. Pericak-Vance and Ren-Hua Chung
• Companies: University of Miami Miller School of Medicine and University of Miami Miller School of Medicine and University of Miami Miller School of Medicine and University of Miami Miller School of Medicine and University of Miami Miller School of Medicine
• Keywords: Autism ; Statistical Genetics ; Pathway Analysis

Abstract

Pathway analysis is useful for identifying variants with modest effects, clustered in multiple genes in a pathway and for studying the relationship of genes in the same pathway for disease susceptibility. Studying genes in a pathway in association with the disease susceptibility extends the use of GWAS data, where single-marker analyses haven't been very successful at identifying the causal variants in our autism study. We developed a novel family-based pathway analysis tool, Pathway-PDT (Abstract 302938, JSM 2011), which uses raw genotype data and accounts for pedigree information in the statistics. When genotype data in families are available, Pathway-PDT can have more power than methods using p-values only. We applied Pathway-PDT to two independent autism GWAS family datasets provided by Hussman Institute for Human Genomics (HIHG) and Autism Genetic Resource Exchange (AGRE), testing pre-defined pathways from the Reactome, KEGG and Gene Ontology (GO) databases. The most significant pathway, the bicarbonate transport pathway (GO: 15701) showed nominally significant associations in the independent analyses (HIHG, P=0.049; AGRE, P=0.009) and in the combined analysis (P< 0.0002).