Read junctions and trim off leading and trailing residues

[torognes]

In many situations there is only junction_aa available in one file, and cdr3_aa in the other. Apart from how to specify the interface, would it be very complicated/give a lot of overhead to have some functionality that allows reading in junctions (with the leading and trailing amino acids) as cdr3s? That would mean: in the case of amino acid, remove the first and last. In the case of nucleotide, remove the first 3 and last 3 nucleotides. We can presume here that no additional checking needs to be done on if the removed aa’s are actually the conserved C and W/F residues (or maybe only check the first sequence). Need to determine user parameters (like how to specify this elegantly for a mix of cdr3/junction files).

[LonnekeScheffer]

In terms of user options, I think this should just be 1 new flag --junction-as-cdr3. Enabling this flag will ensure all sequences from all files are either read in from the cdr3(_aa) columns as is, or by reading the junction(_aa) columns and stripping the leading/trailing amino acids.

Which column to use is determined individually for each file. Some files may only contain cdr3(_aa) and others only junction(_aa). If both columns are present in a file:

• By default, first check if junction(_aa) exists (& is not empty) first, and try cdr3(_aa) second.
• If the --cdr3 flag is enabled in addition, try the cdr3(_aa) column first, and junction(_aa) after.