Erroneous 'region_not_covered' in v4.7.0

[pursellta]

When I use a custom IG library and run the following command using v4.7.0, I get erroneous 'region_not_covered' outputs for the majority of sequences in the 'nMutationsVRegion' when I have exported only functional clones:

mixcr analyze generic-amplicon -t 20 \ --floating-left-alignment-boundary \ --floating-right-alignment-boundary C \ --rna \ --species Bos_taurus \ --library Bostaurus-IGH.json.gz \ --assemble-clonotypes-by VDJRegion \ /home/tpursell/BulkIgFastqs/B18-FLU247-Leader-IGH_S7_L001_R1_001.fastq.gz \ /home/tpursell/BulkIgFastqs/B18-FLU247-Leader-IGH_S7_L001_R2_001.fastq.gz \ /home/tpursell/BulkIgFastqs/mixcr.outs.bota.20260107/bota.B18.FLU247.FSleader

Attached is the alignment report for one of the samples; it shows that 93% of sequences cover the VDJregion so I should not have the error exporting the mutations for the VRegion.

bota.B18.FLU247.FSleaderq.align.report.txt

[mizraelson]

Try using VRegionTrimmed. VRegion requires the full V gene sequence to be present but the last nucleotides are often trimmed in the process of VDJ Recombination.

[pursellta]

I would like to add that I do not have this same error with IGL data processed in the same manner with v4.6.0. I have tried to go back and export the clones with the feature you suggest and the following error:
-aaFeature: Can't construct amino acids for VRegionTrimmed

[PoslavskySV]

@mizraelson

[mizraelson]

Yes, you can only export the nucleotide sequence for VRegionTrimmed because it may or may not be divisible by 3. If you want to export the amino acid sequence, it is better to use {FR1Begin:FR3End}, which is always in frame.

[pursellta]

This fixed the output such that the -aaMutations -aaMutationsRate have an actual output for over 99% of sequences. Do you have any idea why the VRegion which is supposed to be an equivalent gene feature to {FR1Begin:FR3End} does not work but the {FR1Begin:FR3End} does?

[mizraelson]

Yes, because VRegion is {FR1Begin:VEnd}. So it includes the CDR3 part of the V gene, but this part often gets randomly trimmed, so in the majority of actual clones it won't see the VEnd while the FR3End is always present.