PMC:4157143 / 31529-33316 JSONTXT

{"target":"https://pubannotation.org/docs/sourcedb/PMC/sourceid/4157143","sourcedb":"PMC","sourceid":"4157143","source_url":"https://www.ncbi.nlm.nih.gov/pmc/4157143","text":"Genome and transcriptome data from a single large family allowed us to test the utility of various noncoding annotations for predicting the impact of noncoding variants on expression. Here, our goal was to identify those annotations that could inform a functional variant from genome sequence alone. We chose to include the following as potentially informative annotations: ENCODE TF binding, DNase I hypersensitivity peaks, evolutionary constraint, motif disruption as computed by HaploReg, and distance to the TSS. We identified that each noncoding annotation was more informative for predicting the impact of rare variants than the impact of common variants on expression (Figure 5A; Table S7). We observed that evolutionary constraint and distance to the TSS were the most informative for rare variants, and they further increased their utility with increasing strength of constraint and shorter distances, respectively. One potential concern we identified is that we might be only predicting a gene’s ability to harbor an eQTL such that having a rare variant possessing specific annotation might indirectly inform genes tolerant of arbitrary functional variants (both common and rare). However, when assessing whether genes containing different annotations for rare variants were also more likely to have common eQTLs in the population, we saw no significant difference (Figure 5A, right panel). This demonstrates that particular species of rare noncoding variants might be interpretable from genome sequence data alone provided that there is sufficiently high-confidence genotyping of those rare variants. Furthermore, provided increasing availability of genome-interpretation methods, this method offers a means of determining and calibrating the efficacy of different approaches.","tracks":[]}