Add CLI command mergeFasta to combine variant peptides called from different samples

[zhuchcn]

I wonder about the TMT part, cuz that would create variant interactions that should not be present in any sample. I havn't tried this but it sounds safer to do sample specific databases, and then do a "combineFASTA" to create a database for the entire plex? sounds like a to-do for MPG

Originally posted by @lydiayliu in uclahs-cds/pipeline-call-NonCanonicalPeptide#4 (comment)