Up to now, I've been aligning my RNAseq PE samples to a reference whole genome build from Ensembl/gencode. My work involves calling somatic variants at specific regions on chr1, 8, 11, 12, 17, 22.
I have a BED file of my genes of interest, but is it preferred to build a custom reference genome based on the BED file or to concatenate chr1, 8, 11, 12, 17, 22? If using BED is preferred, would I use
fastafrombed function in bedtools?
I understand that my custom reference genome will not have unplaced contigs included but I'm curious how bad this will be if I'm only concerned with calling variants on known regions in the first place.
Curious to hear your thoughts and input.