0
$\begingroup$

I'm trying to map ONT long reads to a portion of a gene I'm looking at. The region is about 25bp long. When I search for the region in the document it pulls up the sequence in every read but when I map to reference with minimap2 it produces no contigs.

Is 25bp too small to be a reference? I can potentially extend it but it's right in an area of conservation in this region of a gene and I would rather not extend much farther.

$\endgroup$
2
  • 2
    $\begingroup$ Can you explain why you would want such a small sequence? If you're mapping, why not use the entire genome? I don't know the specific step size used by minimap2 but I would be very surprised if 25bp is enough, yes. Can you explain your use case in a bit more detail? Why would you want to use a reference that is shorter than even short reads, let alone long reads? What is the perceived benefit? Why is it that you would rather not extend? $\endgroup$
    – terdon
    Mar 15 at 12:47
  • 1
    $\begingroup$ If you don't have a reference genome for your organism/target, you should assemble the reads instead of trying to align them to a short conserved area. I think de novo assemblers exist that are developed specifically for Nanopore data. $\endgroup$
    – Cloudberry
    Mar 15 at 20:32

1 Answer 1

3
$\begingroup$

That's not the right way to do it.

You either have reference or not. If not, you need to assemble one (that problem is called genome assembly). If yes, you map all the reads to the whole reference - the process of mapping is made for this - placing individual WGS reads to their corresponding locations and as a matter of fact, missing portions of the reference can mess it up (that was one of the reasons why there was a push for T2T for human).

Once the reads are mapped, you can use any genome browser (e.g. IGV) to zoom in to the exact location you are interested in and check the stack of reads mapping in there. Given how small the region is, manual inspection of the quality of the mapping is the best you got.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.