Saturday, February 18, 2017
Exhibit Hall (Hynes Convention Center)
Brian Abraham, Whitehead Institute for Biomedical Research, Cambridge, MA
Genome-wide association studies have catalogued thousands of genetic variants that are statistically linked to human diseases. Most disease-associated variants fall within transcriptional regulatory elements called enhancers and may thus perturb the binding of transcription factors, causing dysregulation of disease-relevant genes. Only a few such functional variants have been described because of the absence of a systematic approach to identify this class of variant. Here we describe a method to identify disease-associated variants that affect transcription factor binding, and assign them to appropriate disease genes, by exploiting recent insights from enhancer biology and genome structure. We identified disease-associated DNA variants predicted to alter the binding of transcription factors of disease-relevant cells to enhancers. These were assigned to target genes by using new maps of Insulated Neighborhoods, which are chromosomal DNA loops that constrain the ability of enhancers to act on genes located only within the neighborhood. The use of Insulated Neighborhood structure to predict the target genes of enhancers revised the majority of proximity-based assignments of variant-containing enhancers to genes. For selected variants, experimental analysis verified that transcription factor-binding was altered and the target gene dysregulated. The systematic approach we describe here provides a new predictive guide to functional enhancer variants and their target genes, and may thus accelerate the discovery of the transcriptional mechanisms that contribute to disease and perhaps new therapeutic development for these diseases.