Novel approaches for detecting and analyzing stuctural variants in personal genomes

点击翻译按钮获取中文摘要

Genomic structural variants (SVs), such as copy-number variants or large balanced inversions, represent a major form of genetic variation with an impact on the human genome that is similar to single nucleotide polymorphisms (SNPs). However, compared to SNPs, our understanding of SVs is limited. The resolution of published surveys has thus far been insufficient for mapping the start- and end-points (i.e., breakpoints) of SVs, hampering detailed analyses of SVs. Here we propose to study SVs in unprecedented detail: we will [1] develop computational approaches to identify SVs and analyze their extent in the genome; [2] mine published datasets to unravel SV de novo formation processes; and [3] apply approaches to measure SV de novo formation rates and to infer the influence of natural selection. Specifically, first we will systematically collect information on SVs with known breakpoints and standardize their description in a library. Scanning short DNA sequencing reads against the library will enable accurate SV-detection and thus add significant value to “personal genomes” that presently lack detailed SV analyses. Second, we will develop a framework for untangling SV formation by devising computational approaches for inferring the likely ancestral state (by comparison with primate genomes) and the likely causal mutational mechanism (by breakpoint analysis) at each SV locus. Third, we will estimate SV formation rates experimentally and SVoccurrence frequencies computationally to assess the factors contributing to the abundance and distribution of SVs in the genome.