QIAseq Targeted RNA Panel TCR Library Kit align report

[ejc043]

Hello! Thank you so much for this amazing package.
We are currently working with data from QIAseq Targeted RNA Panel TCR Library Kit and I'm analyzing it using v4.7.0.
Using this thread, I analyzed our data using this script:

#!/bin/bash

sample=NM_
r1=$(ls ${sample}*R1*.fastq.gz)
r2=$(ls ${sample}*R2*.fastq.gz)

echo $r1 
echo $r2
mixcr analyze qiagen-mouse-rna-tcr-umi-qiaseq \
    --tag-pattern "^(R1:*)\^(UMI:N{18})N{10}(R2:*)" \
    ${r1}  \
    ${r2} \
    ${sample}

This is the QC output:

 Successfully aligned reads:                           86.67% [OK]
  Off target (non TCR/IG) reads:                        10.51% [WARN]
  Reads with no V or J hits:                            1.17%  [OK]
  Reads with no barcode:                                0.0%   [OK]
  Overlapped paired-end reads:                          90.86% [ALERT]
  Alignments that do not cover CDR2_TO_FR4:             0.75%  [OK]
  Tag groups that do not cover CDR2_TO_FR4:             0.12%  [OK]
  Barcode collisions in clonotype assembly:             0.15%  [OK]
  Unassigned alignments in clonotype assembly:          1.83%  [OK]
  Reads used in clonotypes:                             97.82% [OK]
  Alignments dropped due to low sequence quality:       0.0%   [OK]
  Clones dropped in post-filtering:                     0.0%   [OK]
  Alignments dropped in clones post-filtering:          0.0%   [OK]
  Reads dropped in tags error correction and filtering: 0.82%  [OK]
  UMIs artificial diversity eliminated:                 69.99% [ALERT]
  Reads dropped in UMI error correction and whitelist:  0.0%   [OK]
  Reads dropped in tags filtering:                      0.82%  [OK]

this output is pretty consistent across all our samples. How could the altert for overlapped pair-end reads and UMIs artificial diversity eliminated alert be explained?
Thank you so much!

[mizraelson]

Overall, the results look pretty good, and I wouldn’t worry. It’s possible that some reads don’t overlap due to long inserts, but the percentage you’re seeing is acceptable. A high UMIs artificial diversity eliminated value is also common and typically results from sequencing errors. The important part is the line Reads dropped in tags error correction and filtering: 0.82%, which means that less than 1% of reads were filtered out and the correction went well —this is very good.

[ejc043]

This is so helpful— thank you!