search

jts / nanopolish

Signal-level algorithms for MinION data
MIT License
566 stars 159 forks source link

Mapping Quality option #931

Closed vahidAK closed 3 years ago

vahidAK commented 3 years ago

Hi @jts, I hope you are doing well,

In the newer versions of nanopolish, you have implemented a mapping quality of more than 20 during methylation calling. But, as far as I am aware, there is no option in the tool to specify mapping quality in case we were interested in other low-quality mapped reads. Comparing to older versions it results in detecting about 1-2M less CGs. Does it mean those CGs called by older versions are not reliable at all? Do you have any plan to add an option to specify mapping quality for the tool during methylation calling?

Many thanks, Vahid

jts commented 3 years ago

Hi @vahidAK,

Comparing to older versions it results in detecting about 1-2M less CGs. Does it mean those CGs called by older versions are not reliable at all?

I haven't made this comparison, and I think it would be difficult to do so as if a nanopore read has low mapping quality it is unlikely that the bisulfite data we'd compare to maps well. So it is hard to say whether the calls are reliable or not.

Do you have any plan to add an option to specify mapping quality for the tool during methylation calling?

Sure, I can make this an option. I'll leave this issue open until I do so.

Jared

vahidAK commented 3 years ago

Thanks a lot, @jts

It would be great if you add the option

jts commented 3 years ago

I just added this in the latest commit, give it a try