Indels detected only in regions where forward and reverse reads overlap in matched normal libraries

[scddfmsku]

Hello, I am currently using nanoseq for our basic biology project, and I really appreciate the functionality it provides. I have a question regarding indel detection.

We prepared nanoseq and matched normal libraries from genomic DNA digested with the restriction enzyme, performed sequencing (150 bp paired end reads), and analyzed base substitutions and indels using the nanoseq analysis pipeline.

When we visually inspected the indels reported by the nanoseq analysis pipeline using the IGV genome browser, we found that all of the indels were observed only in regions where the forward (plus) and reverse (minus) reads overlapped in the matched normal libraries (but not the nanoseq libraries), that is, where the genomic insert size was short, typically less than 300 bp. In contrast, when using a standard whole genome sequencing (WGS) library for the matched normal sample (prepared by sonication of genomic DNA), this limitation was not observed, and indels were detected even in regions where the forward (plus) and reverse (minus) reads did not overlap in the matched normal libraries.

We are currently uncertain whether there is an issue with our analysis or interpretation, or if the results described above represent the pipeline’s default behavior. If the latter is the case, we would like to know whether there is a way to modify the parameters or source code so that indels can be detected in both overlapping and non-overlapping regions of the forward (plus) and reverse (minus) reads in the matched normal libraries.

We believe that, in the case of base substitutions, this kind of behavior can be controlled by the -b option of var.

I understand you are very busy, but we would be sincerely grateful for any advice you could offer. Thank you for your time and consideration.

[fa8sanger]

Hi, I understand you created two matched normals, one with enzymatic restriction (nanoseq-style, or "RE") and one with standard WG sequencing. When using RE matched normals the "-b" option should always be 0 because very often coverage has only forward or reverse reads. For standard WG matched normals you can set -b > 0 but we don't see much benefit in that.

I've never seen the pattern you mention. That fwd and rev overlap is dictated by the distance between the two restriction sites. If they are less than 300 bp apart, with 150 bp reads you'll get some overlap.

[scddfmsku]

Thank you very much for your prompt response. I truly appreciate it. I am embarrassed and sincerely apologize for my delayed reply in comparison...

As you correctly pointed out, we are indeed working with two matched normals: one generated using enzymatic restriction ("RE") and another from standard whole genome sequencing ("WGS").

We realized some time ago that, when using RE matched normals, the -b option for "runNanoSeq.py var" should always be set to 0, as you've noted. However, for "runNanoSeq.py indel", since the -b option does not seem to be available, we have been uncertain about the best approach and have been considering how to handle this.

We will take another careful look at our data in light of your insights. Thank you again for taking the time to guide us. Your support is greatly appreciated.

Best regards,