Clarification on --variants parameter for A-to-I editing detection

[baiyin051]

Hi REDItools3 developers,

Thank you for providing this great tool!

I have a quick question regarding the --variants parameter in reditools analyze. If I want to specifically detect A-to-I RNA editing events, which are typically observed as A>G (or T>C) mismatches, should I use:
--variants AG TC
or
--variants AG CT
I noticed that the default is --variants AGCT, but it’s not entirely clear what that means in the context of detecting only canonical A-to-I edits. Could you please clarify?

Thanks in advance for your help!

[ahanden]

If you are interested in detecting A-to-I editing, the value you choose for the --variants parameter would depend on whether you have single or paired end library. If it's single end, then you would be correct with --variants AG TC. For paired end libraries, REDItools should be able to determine the strandedness of the reads and will convert any anti-sense results to their complement. In that case, you should only need --variants AG.

The default value chosen for --variants is actually for detecting two kinds of editing: A-to-I and C-to-U (the former performed by ADAR1 and the latter by APOB). But this isn't explicitly stated anywhere in the tool's documentation. I think it would be a good idea if we included a section in the readme for choosing the value for --variants.