How is `cid_full_length` assigned?

[ejohnson643]

So I've been trying to understand how the _report.tsv is generated from the _barcode_report.tsv and the _cdr3.out files in the trust-simplerep.pl script and I believe that I have correctly determined that the CDR3s counted in the _report.tsv are aggregated across barcodes using the V-D-J-C-CDRnt annotations as a unique key. This is all fine, but I was checking whether there were CDR3s in the report.tsv that have the same V-D-J-C and CDR amino acid annotations, to see how different the different CDR3s in the file are. As an example: So we can see that there are 5 different CDR3s with identical V, J, and CDRaa, but which differ in the exact nucleotide sequences. However, when I look at whether these CDR3s are "full length," only the first one is, despite them all having nearly identical CDR3 nucleotide sequences: Could someone provide some insight into how exactly the "full length" determination is made and how it could show up differently in these elements of thereport.tsv? Thank you!

[mourisl]

The full length means the underlying contig contains the full-length receptor variable domain: 5' of V gene to the 3' of J gene. It is more strict than the complete CDR3.

[ejohnson643]

Thanks for your help!

To make sure I understand: cid_full_length is a property of the contig, not necessarily of the CDR3.

The idea behind indicating this information in the _report.tsv is that it let's the user know whether the CDR3 was generated from a contig that contains the ends of the V and J gene?

[mourisl]

Exactly.