I'm using igblastn 1.4.0 with TRAV and TRAJ databases from IMGT. It aligns the sequences in the query correctly (the FR3 sequence is well-identified) but it fails to set the right bounds for CDR3.

This is one example from an aligned sequence to the germline:

<-------------FR3-IMGT------------->                                              
 S  A  Q  L  S  D  S  A  L  Y  Y  C  A  L  S  S  N  A  V  A  K  L  T  F  G  A  G
TCAGCGCAGCTGTCAGACTCTGCCCTGTACTACTGTGCTCTGAGTTCTAATGCCGTCGCTAAGCTCACATTCGGAGCAGGA

As you can see igblastn correctly identifies the FR3 region up the Cys amino acid. Just afterwards the CDR3 starts, but igblastn tells me this:

CDR3-IMGT (germline)    37      45      9

That is, the CDR3 sequence is only 9 bases long, GCTCTGAGT or ALS. However, CDR3 should end with either a F or a W amino acid (not included). In this case it would be F, with CDR3 being: GCTCTGAGTTCTAATGCCGTCGCTAAGCTCACA.

How can I tell igblastn to do this? Is it configurable somehow?

Right now I pass the following arguments:

-germline_db_V mouse_f_orf_inframeP_TRAV_blast-edited.fasta
-germline_db_J mouse_f_orf_inframeP_TRAJ_blast-edited.fasta
-germline_db_D mouse_f_orf_inframeP_TRBD_blast-edited.fasta
-organism mouse
-domain_system imgt
-query myquery.fasta
-ig_seqtype TCR
-auxiliary_data optional_file/mouse_gl.aux
-show_translation -outfmt 3

I can't see anything wrong here. Any help would be appreciated.