Definition of productive clones

[rvalieris]

Hello,

I'm trying to replicate the productive column from the airr format with the exportClones command.

exportClones have a -isProductive <geneFeature> flag, which gene feature should I use here to have the same output as the exportAirr command ?

from what I tested it seems to be CDR3, but I wanted to confirm this, and then also ask what the flag --export-productive-clones-only is doing because that gives me different results.

[mizraelson]

Hi,
Yes, the correct feature is CDR3. Can you give an example where using --export-productive-clones-only gives a different result?

[rvalieris]

oh I see, I might be doing something wrong then, here is what I tried:

PREFIX=SRR4026007 # https://ncbi.nlm.nih.gov/sra/?term=SRR4026007
R1="$PREFIX"_1.fastq.gz
R2="$PREFIX"_2.fastq.gz

mixcr -Xmx10G analyze generic-amplicon-with-umi \
  --limit-input 10000 \
  --no-json-reports \
  --species hs \
  --rna \
  --tag-pattern '^(R1:*) \ ^(UMI:N{17})(R2:*)' \
  --floating-left-alignment-boundary \
  --floating-right-alignment-boundary C \
  --assemble-clonotypes-by '{CDR1Begin:FR4End}' \
  $R1 $R2 $PREFIX

mixcr exportClones --chains IGH -isProductive CDR3 -hasStops CDR3 -isOOF CDR3 $PREFIX.clns test1.tsv
mixcr exportClones --export-productive-clones-only --chains IGH -isProductive CDR3 -hasStops CDR3 -isOOF CDR3 $PREFIX.clns test2.tsv

then I count how many clones are in each file, according to the isProductive column:

$ cut -f35 test1.tsv | sort | uniq -c
     22 false
      1 isProductiveCDR3
    100 true

$ cut -f35 test2.tsv | sort | uniq -c
      1 isProductiveCDR3
     91 true

$ cut -f36 test1.tsv | sort | uniq -c
    122 false
      1 hasStopsInCDR3

$ cut -f36 test2.tsv | sort | uniq -c
     91 false
      1 hasStopsInCDR3

$ cut -f37 test1.tsv | sort | uniq -c
    100 false
      1 isOOFCDR3
     21 region_not_covered
      1 true

$ cut -f37 test2.tsv | sort | uniq -c
     91 false
      1 isOOFCDR3

I expected test2 would have then same 100 productive clones as in test1.

[mizraelson]

Hi,

Yes, the --export-productive-clones-only option examines the entire region, not solely the 'CDR3'. Typically, we anticipate that stop codons or frame shifts would occur exclusively within the CDR3 region and the outputs are the sane. However, in the dataset you shared (SRR4026007), some clones have hypermutations resulting in stop codons appearing in regions outside of CDR3. Hence the differences.

[rvalieris]

the --export-productive-clones-only option examines the entire region, not solely the 'CDR3'

ah ok, that makes sense, thanks.