Using Stochastic Search Gene Suggestion to Identify Single Nucleotide Polymorphisms Associated with Childhood Leukemia Risk in Case-Parent Triads

• Activity Number: 311
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Michael Swartz*+ and Ying Cao and Darryl Nousonme and Philip Lupo and Michael Scheurer
• Companies: The University of Texas Health Science Center At Houston, School of Public Health and The University of Texas Health Science Center of Houston, School of Public Health and University of Southern California and Baylor College of Medicine and Baylor College of Medicine
• Keywords: Stochastic Search Variable Selection ; Candidate Gene Approach ; Single Nucleotide Polymorphism ; Case-Parent Triads ; Acute Lymphoblastic Leukemia

Abstract

Stochastic Search Gene Suggestion (SSGS) is a Bayesian variable selection procedure used for gene mapping that incorporates genetic structure in hierarchical priors for model selection. Specifically, The selection priors model selecting alleles within genetic loci (Swartz et al, 2006, Biometrics). For this project, we updated SSGS to analyze case-parent triads using single nucleotide polymorphisms (SNPs) by organizing the hierarchical priors to model selecting SNPs within genes. We also used the priors to incorporate linkage disequilibrium between loci. We applied these selection priors to the conditional logistic regression likelihood to model genetic transmission within case-parent triads. We then utilized our SSGS for SNPs method to identify risk alleles among 118 case-parent triads with childhood acute lymphoblastic leukemia. For this project, we interrogated 128 tagging SNPs in 16 genes in the folate metabolism pathway. Our approach identified two SNPs with a Bayes Factor >1, one with a Bayes Factor >7. Based on our findings, we believe the SSGS approach for gene mapping is a useful method in identifying disease associated SNPs among case-parent triads.