2
$\begingroup$

I'm analyzing some Oxford Nanopore sequencing reads, with several articles stating the sequencing error is between 5% and 15%. Since its 1D and not 1D^2 sequencing i would assume the sequencing error is closer to 15%.

I'm using Minimap2 aligner to align the reads to the reference genome, and then im choosing those reads with a mapQ of 60, to analyze.

Where i know with the phred score mapQ of 60 means 99.9999% of the bases are called correctly.

So my question is, if the mapQ is 60, then does it really matter if my sequencing error is 15%? Or how is the sequencing error correlated with mapQ? Becuase i guess you could still obtain a good mapping quality with a few mismatches?

Or am i completely off target.

Thanks for your time and help.

$\endgroup$
  • $\begingroup$ Recent 1D data is also about 95% accurate. $\endgroup$ – Wouter De Coster Sep 17 at 9:26
3
$\begingroup$

Where i know with the phred score mapQ of 60 means 99.9999% of the bases are called correctly.

That's not at all what the MAPQ means, it means that the sequence is very likely to have originated from that place in the genome, whether the underlying sequence is correct is another question all together.

For long nanopore reads, sequence quality is not well correlated with MAPQ. MAPQ will tend to be either very low (0 or 1) or very high (60) since you tend to have kb+ reads and those will largely be uniquely assignable in the genome. The sequencing quality of nanopore reads is only an issue when you care about something at the base level, such as SNP calling or finding exact splice junctions. Even then, you mostly need sufficient coverage to get around this.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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