Splitting or trimming at internal adapters in long-reads.

No issue but an enquiry about the suitability of what I want to do.

I have some long-read data that is essentially concatemers of genomic fragments separated by bridge adapters much like sci-Hi-C but multiple concatamers not pairwise.

I've been trying to find a tool that can ideally split or trim the reads where it finds an adapter multiple times. The tricky part is I do not know where or how many times the adapter is present. I thought about running cutadapt iteratively but how do you know you've finished? https://github.com/marcelm/cutadapt/issues/621#issue-1220110397 Perhaps this feature would provide a list of readIDs that still need to be trimmed.

--untrimmed-output would output the original read without adapter sequences?

From reading the documentation it seems that multiple hits can be permitted but is this compatible with keeping the sequences in between the found adapters?

Another suggestion from someone was to use the action=mask to convert any multiple instances of found adapters to Ns and then trim the Ns, a bit like splicing out introns.

For now, I am just interested in removing the full bridge adapter sequence resulting from ligation, therefore, I supply this sequence as it should appear from 5'->3' in the direction of nanopore sequencing. I include an instance of just one ligation event but potentially reads may have several.

Start by reviewing the documented behavior of multiple adapter hits, --untrimmed-output, and action=mask, along with the related cutadapt issue #621. Determine whether the requested behavior is a new feature or a documentation clarification, including how sequences between repeated adapters should be retained. Done means the supported behavior and a way to identify or process reads with multiple internal adapters are clearly established.