add flexibility for other PCGs

[bwprice]

what settings are currently hard coded for COI and can we move these to config so it can pull out any protein coding gene if specified via config?

[SchistoDan]

Note:

• Only fasta_compare has a hardcoded barcode target step - currently uses coordinates to extract barcoding regions from full consensus sequences (e.g. 40-700bp = COI barcode region). Users can list barcode (e.g. COI) in BGEE config file to change barcoding region for each target.

• Alternative: User supplies forward and reverse primers instead, which are used to run seqkit amplicon for in silico extraction of barcode region.

Open to other suggestions!

[bwprice]

need to consider possible issues with the reference sequences used:

• if they weren't the full gene then coordinates may not be an accurate approach
• if they don't include the primer sequence then primer based extraction will not work (though more flexible than coordinates!)

Options I see:

1. iterate through primer extraction first, if this fails then fallback hard-coded in config
2. align to a single global user supplied reference (AA or DNA format) OR HMM model per target gene

• (2a) align reference to global at the start of run to ensure it covers the correct region before any mapping happens
• (2b) align extracted sequence after assembly and extract the barcode region (assumes reference included this region)