neb-mouse-rna-xcr-umi-nebnext QC ALERT

[Victor-K27]

Hi all,

I am running the neb-mouse-rna-xcr-umi-nebnext analyze preset of the MiXCR package.

I am getting a low number of successfully aligned reads, and a high number of off-target reads.
I am also getting a lot of alignments and tags that do not cover the VDJRegion, and high artificial UMI diversity.
I believe that this indicates weak enrichment for TCR sequences?

Regarding the trimming, I am getting an average of 5-20 nucleotides, which is hopefully ok?

I am attaching bellow a copy of the qc reports. I have been going through them, but some comments from someone more experienced would be very appreciated!

I am also wondering, is it normal to have that high ratio of TRB to TRA reads? In this sample for example, TRA reads represent 10% of the repertoire with TRB reads at 88%.

I am wondering (as far as I was told this pipeline is new), if I am the only one experiencing issues?

alignQc copy.pdf
Spleen_1.qc.txt
Spleen_1.align.report.txt
Spleen_1.assemble.report.txt
Spleen_1.refine.report.txt

Thank you very much for your time!
Victor

[mizraelson]

Hi Victor,
Generally speaking:
• A 30% alignment rate is quite low, and a high number of off-target reads suggests some primer misannealing.
• The refine report (UMI diversity) looks fine. It’s common for many UMIs to be filtered out, but the important thing is that more than 90% of reads were preserved.
• Trimming 5-20 nucleotides should be okay, and we still see that 90% of reads overlap. However, it’s strange that the full VDJ region is only covered in 30% of reads. It’s hard to say more without looking at the data.
• 10% of TRA compared to TRB is quite low and might signify some wet lab issues.

If possible, you can share the raw FASTQ data ([email protected]) for this sample, and I will take a look and probably be able to provide more insights.

[mizraelson]

Hi Victor,
I looked into the non-aligned reads for one of the samples you shared. The majority of the off-target reads align to other genes (e.g., lacZ, Glyr1). Some reads do align to the TCR locus, but the alignment looks like this, for example:

CAAGGAGACCTTGGGTGGAGTCACATTTCTCAGATCCTCTGATGTTCCACATTCACTCCTACACTGAGATGTAAGAGAGCTGTGGTGGCTTGCATAGGAAAGCCGCAATCCTATTACCAAAAGCCTGGTCCCTGAGCCGAAGGTGTAGTCGGAGTTTGCACATCAGAATACAGATACTCGAATATGGACACGGAGGACATGCTTTGTCCGAAGAGAGACCTGGAAATTTACCTGAAAGATATCTTACCTACAACTGTGAGTCTGGTTCCTTTACCAAAGAAGACTTCTGTGTTTGCCCTGT

This read covers two J genes and the DNA in between, which doesn’t resemble a regular rearranged TCR mRNA.

Or this:

CAAGGAGACCTTGGGTGGAGTCACATTTCTCAGATCCTCTGATGTTCCACATTCACTCCTACACTGAGATGTAAGAGAGCTGTGGTGGCTTGCATAGGAAAGCCGCAATCCTGTGGAAGATATTATTCACAGGTTGAACTCTTATTTTTTCAGTGCTCAGTGAACCTCCCTGATGAGTTTCAGAATTCTAGTTTACAACTAAATGTTCTATTCTTTAATATCCCGTAGATCAATGAATTTAGAGTATAAGATCGGAAGAGCACACGTCTAACTCCAGTCACATTACTCGATCTCGTATGCC

This read aligns to the beginning of the C gene but then continues with some DNA sequence from the genome between the J and C genes.

Although these reads align within the locus, they clearly do not represent a complete VDJ rearrangement. The origin is unclear and could be due to DNA contamination or a complicated, incomplete rearrangement.