Unexpected Inserted Sequence Between Repeat Motifs

[gotanjiayu]

Hi Developer,

I’ve been using TRGT for RFC1 repeat detection and encountered an issue I’d like to ask about. In the output VCF, I observed that the reported allele sequences include long stretches of non-RFC1 sequence between the repeat motifs.

This sequence is quite long and doesn't appear to follow the expected repeat motif. In the corresponding plot, this region is visualized as a long gray segment between colored repeat blocks, suggesting it may have been interpreted as a large inserted or interspersed region. check below the TRGT plot.

BUT, when I check the bam in the IGV, I didn't see the large "GAAATGAATGGAATGAAATGAAATGGAATG" sequence between the motifs.

Thanks a lot for your help!

[pbsena]

Hello,
Thank you for reporting this interesting example. The reason the reads are colored in gray is likely because the motif units are novel and are not part of your BED catalog. If you want to give them a color, you might want to re-run TRGT by adding AATGGAA to the MOTIFS field in your RFC1 entry.

I'm not 100% sure on the IGV plot, but is it possible that these sequences are contained in the 2,776 bp insertions present in many of the reads? You can also look for them in the .spanning.bam output by looking for reads that have the TR:Z:RFC1 tag.

We are currently developing a sub-command that allows you to align individual reads to the consensus and visualize them on IGV, and I believe this may be helpful for this use case as you'd be able to see if individual reads contain the motif more easily. If you are interested in testing it I'd be happy to share a pre-release binary with you if you give me an e-mail I can reach out to.

[gotanjiayu]

Thanks for responding quickly. I was wondering if TRGT has any feather for Sequence Error Tolerance: Allows for minor mismatches/indels in motifs? it seems like there're some artifact "T" in the large repeats, so causing the grey area in the TRGT plot.

[pbsena]

We used to, but adding this tolerance causes problems in identifying biologically relevant single base pair interruptions in important expansions, so currently TRGT treats any single bp change in the consensus as an interruption and highlights them in gray