Marcel Martin

I’ll also see whether I can make it so that you can specify `{name2}` as a template variable. That would then explicitly declare that you want to use the matches...

Thanks for reporting! First, can you confirm that Cutadapt creates an index of the 5' primer sequences? You should get a message that says `Building index of *n* adapters ...`....

@eboyden Just briefly, would it be possible to share some of that data? Privately via e-mail if you want. I could create a simulated dataset if not.

> Unfortunately I don't think I can share any data. No problem. Can you let me know which command-line options you used? (Of course feel free to omit sequences or...

Thanks. I’m busy with other things, but want to look into this as soon as possible. In the meantime, maybe you could do an experiment: You could increase the maximum...

> Sorry for the confusion: v3.8 doesn't support the `^file` nomenclature, so in cases where I didn't use it, I used a file of anchored sequences instead; I assume it...

There is no built-in way to set a variable error tolerance for a single adapter like this. Also, the alignment between the read and the adapter is not computed, so...

I’ve done a couple of optimizations and was able to make indexing when `--no-indels` is used faster by a factor 1.75 (17 s instead of 30 s for a set...

And BTW I did not note any significant contribution to runtime from the k-mer heuristic, so I don’t think we need to make any changes there.

> A few reads in the NTC contain Ns at the beginning, but it's If the time per read is based on the wall clock time, isn't it possible that...