Using other reference genome builds

[mhSepehri]

Is there any way to use reference genomes of organisms other than hg19, hg38, mm9, and mm10?

I am trying to use facets for dog samples (canFam3) and the first part of generating pileup file works fine and creates an output with all 38 + x chromosomes output.csv.gz, but it cannot produce final results correctly.

[dariober]

Is there any way to use reference genomes of organisms other than hg19, hg38, mm9, and mm10?

Hi- I'm afraid this is not possible at the moment and implementing it has been on my todo list for a while - it shouldn't be too difficult. facets (the R package) supports it, though.

---

Note to myself - this is how it could be implemented:

• --gbuild/-g option should accept also a path to a bed file where the 4th column is the percentage GC in 1kb windows (users can use bedtools to prepare this file). Make a list (gcData) from this file where items of the list are vectors of %CG (4th column). This the format in the pctGCdata package.

• If gbuild is a bed file, function make_header can get the chrom sizes from gcData.

• In preProcSample function call, if gbuild is a bed file set set gbuild="udef", ugcpct=gcData. Use ugcpct=NULL otherwise.

• Function reset_chroms can probably do nothing if gbuild is a bed file.

Alternatively:

• If --gbuild is a preset string, rename chromosomes as appropriate and assign the pctGCdata list to object gcData. If gbuild is a bed file, read it, make it a list of chroms and assign it to gcData.

• Proceed as if gbuild is always a custom genome and use gcData instead.