Restricting str-profile generation to contigs/regions provided in .bed or similar

[am8265]

Hello developers,

Is it possible to add the feature of restricting to specific contigs/regions provided in .bed/similar while generating str-profile
using this command -
ExpansionHunterDenovo profile

Hopefully this will be useful for reducing runtime for bad incomplete contigs.

Thanks!

[egor-dolzhenko]

Thanks for the suggestion! The main hurdle with implementing this is that the reads originating in expanded repeats on one contig often get misaligned to other contigs. To avoid jumping around a BAM file and recovering these reads, EHdn just streams through the entire BAM file. This usually ends up being much faster than a targeted analysis of a large number of regions.

However, if you are interested in profiling just a small number of regions, we could consider implementing a separate targeted workflow.