Pipeline for linking genomic variants to affected features

In order to better (faster) query for affected genes (and other genomic features) due to mutation (genomic variation), we will need a regular automated processing pipeline that will create tables of variation-effects. It would take all genomic variants (SNVs, CNVs, etc.) in a single organism, and scan for genomic coordinate intersections with all known genomic features on both strands (not just genes, but also things like enhancers, histone binding sites, etc.) Because we will be slurping in new annotation data throughout the project, we would have to regenerate this on a regular basis. There might need to be some kind of fancy processing done if we want to include known targets of enhancer/insulator regions (since they can often be long-distance and affect multiple genes).

I think precomputing the intersection of the affected genes will make the other lookups much faster. Then, downstream, the user can select which feature-relationship types they want to include in their searches. For example, they might want to only examine mutant phenotypes derived from regulatory regions of specific gene classes.