A posterior probability of linkage & association study of 111 autism candidate genes

Autism is a neurodevelopmental disorder with a complex genetic basis. In this study we investigated the possible involvement of 111 candidate genes in autism by studying 386 patient families from the Autism Genetic Resource Exchange (AGRE). These genes were selected based on their functions that relate to the neurotransmission or central developmental system. In phase 1 of the study, 1497 tagSNPs were selected to efficiently capture the haplotype information of each gene and were genotyped in 265 AGRE nuclear families. The cleaned genotype data were analyzed through the Kelvin program to compute values of Posterior Probability of Linkage (PPL) and Posterior Probability of LD given linkage (PPLD), which directly measure the probability of linkage and/or association. Consistent supportive evidence for linkage was observed for EPHB6-EPHA1 locus at the 7q34 region by two- and multi-point PPL analysis. Some evidence for association was obtained from the intronic SNP rs2242601 of the EPHA1 gene (PPLD = 10.4%), and multiple SNPs from the MECP2 gene at Xq28 (PPLD range from 5∼9%). Using a subset of the newly released AGRE genotype data from the Affymetrix 5.0 high-density SNP array, further evidence for association was obtained for 6 markers located 90kb distal of EPHA1 gene (PPLD range from 21% to 40%).;In phase 2 of this study, in an attempt to conduct fine mapping as well as to replicate our phase 1 results in a set of 123 additional AGRE family samples, additional SNPs were selected from the EPHA1 and MECP2 gene region for fine-scale analysis. Strong support of association with autism was observed for the markers downstream of the EPHA1 gene using the original families, with the SNP rs7801889 showing a high PPLD value of 62%. Markers from the MECP2 gene region remained moderately associated with PPLD values around 8%. Nonetheless, none of the SNPs showed any support for association in the additional family samples. These mixed preliminary results suggested the polymorphisms within and downstream of the Ephrin receptor A1 gene as potential novel susceptibility loci for autism. Limited support for the role of MECP2 in autism etiology was also observed.