nanoporetech
Methylation analysis in plants
Hi ONT Dorado Team. I am planning to do a methylome sequencing in a couple of grass species and was wondering if you have benchmarks that you can share on the accuracy of Dorado for plant genomes in terms of basecalling and 5mC methylation calling? If you have not tested Dorado on plants, is it safe to assume that I can also expect Q20 and >99% methylation accuracy? I remember Guppy having plant specific basecaller before but I am not sure if this is also the case with Dorado. Thanks!
Our production modified base caller calls both 5mC and 5hmC in all-contexts. Our latest accuracy results on "all-contexts" (variety of non-CG contexts including reference sequences from some plant species) can be found in the most recent modified bases update talk from NCM in December (video link here). The primary slide of interest is copied below. We did not show the collapsed 5mC-only accuracy in all-contexts, but you can see from the second slide below that a significant improvement is expected when dropping 5hmC calls in CG-contexts. The modified base collapsing functionality is part of our modkit tool.
@EpiAllele does the reply from Marcus answer your questions and are there any more questions or can we close this issue?
Kind regards, Rich
Hi @HalfPhoton sorry for the delay. Yes, the answer from Marcus is what I am looking for in terms of methylation accuracy, and thanks for the link to the NCM video. Very informative.
However, I am still looking for benchmarks/information about basecalling accuracy in general (for canonical bases) in plants. I only hear some anecdotes about nanopore sequencing in plants not really being able to attain the Q20 modal accuracy. The publications that I am seeing so far have used the old R9 flow cell and chemistry. I was wondering if you have done some internal testing on plant whole genome sequencing and what can we expect in terms of accuracy.