Despite the existence of a core set of features and affected biological networks, autism spectrum disorders (ASDs) are genetically heterogeneous. While variants in hundreds of genes have been implicated as causal or risk-conferring for ASD, a large percentage of the heritability of ASDs remains unaccounted for. This suggests that a number of inherited causal or risk mutations linked to autism have gone undiagnosed.
Production of functional proteins from their coding genes requires that transcriptional and post-transcriptional processes, such as splicing, occur properly. During splicing, more than 90 percent of the material in an initial gene transcript is removed in the form of intervening, noncoding segments (i.e., introns). The sequences required for the correct recognition of these segments present targets for disease-related mutations. Indeed, results from William Fairbrother’s research group at Brown University suggest that at least one-third of all hereditary disease alleles disrupt splicing.
Over the past decade, Fairbrother’s team has developed computational methods and high-throughput assays to detect gene variants that alter splicing. The team proposes to use these computational methods and high-throughput assays to prioritize ASD gene candidates emerging from ongoing sequencing studies and to map splicing elements in ASD risk genes. Results from these studies will be integrated into a software package intended for use in analyzing deep sequencing data to predict genetic variants highly likely to be causal to, or predisposing of, ASDs.