m6anet results compared to modkit

[esh97]

Thanks for updating m6anet to work with the new RNA004 chemistry. I've had luck in the past used m6anet on my RNA002 data sets, but I've now switched to the RNA004 chemistry. When I run m6anet and ONT's modkit on the same dataset, I get differing results. Namely, I'm looking at nuclear and cytoplasmic fractions and the two programs yield opposite results as to where m6A is enriched. Have you or any collaborators compared the output of m6anet to modkit?

Thanks!

[yuukiiwa]

Hi @esh97,

Sorry for the delayed reply!

Internally, I have compared m6anet to dorado+modkit m6A results of RNA004 HEK293T with m6ACE as ground truth, where the overlapping sites detected by the two methods give comparable performance. This being said, I have not dug into the individual sites to look at them.

Thanks!

Best wishes,
Yuk Kei