Could we use Nanopore Q20 reads to construct the meryl dbs?

[zmz1988]

Hi, thanks a lot for the very nice tool! I have a question about building the meryl dbs with the Nanopore Q20 reads. What do you think if I use the Nanopore Q20 reads (only reads QV>20) plus illumina reads to build the meryl dbs?

I know it was not recommended for the previous Nanopore reads, which have not so high accuracy. I'm also aware that pacbio hifi still has higher QV than nanopore Q20 reads though 1/3 of the hifi reads has only 30>QV>20. So I'm thinking if we use the pacbio hifi reads for meryl dbs, then probably we could use the nanopore reads with QV>Q20 (which is a bit less than half the total reads) for meryl dbs?

Thanks a lot!

[arangrhie]

Hello,

This is something not so easy to provide a deterministic answer, because Nanopore reads are so frequently updated. If Nanopore reads are the only option, you could probably try that, with the caveat that Merqury QV is more of an indicative of how well the assembly reflects bases in the Nanopore reads. Any bias or sequencing error remaining in the Nanopore reads carried over to the assembly will not be detected, thus inflating the QV. I'm keeping an eye on the Nanopore R10 duplex reads, which the accuracy is getting higher, however I'd be more hesitant to use kmers from R9 or previous reads, or simplex, even if it is > Q20.

Since you have Illumina reads, why not try first using that alone?

Arang

[zmz1988]

Thanks for replying me! Yes, I already used Illumina, but the QV is not so high, as the illumina reads are from different individuals (the same line same inbreeding species) with the one used for making genome assembly (from Nanopore reads). So thought whether we could include the high-quality Nanopore sequences and see whether the QV increase. But would like to hear the opinion from the experts before we do that. :)