What does modkit extract calls do with reads that do not have a considered motif?

[billytcl]

I think I know the answer to this, but what does extract calls do for reads that do not have the motif of interest?

Eg if I run modkit extract calls --cpg what happens if I have reads with no CG motif? Does it skip over the read and not report it? Do I have to check the modkit log file to see how many were skipped because of this reason? Is the correct way then to not use --cpg (and also call dorado using 5mC instead of 5mCG) if I want to grab all reads in that table?

[ArtRand]

Hello @billytcl, sorry about the delay.

You're correct that if a read doesn't have any basecalled CpGs and you've run the CG model there won't be any rows output in the table. The number of records where zero modified positions are found will be tabulated in the "skipped records", keep in mind that this count also includes non-primary and unmapped records (when using --cpg or --motif). You're also correct that if you want to collect rows where the basecalls are CH but they are aligned to reference CpGs, use the all-context cytosine model and --cpg.