Read splitting

[guangzhaocs]

Previously, I reported a question (get more reads from Dorado). The old question here: https://github.com/nanoporetech/dorado/issues/701

So, could I prevent the splitting using some parameters (e.g., --no-trim)? Or how to combine these splitting reads to get the full output corresponding to the original reads? (my data is RNA)

I do not find the introduction of the the read splitting of concatemers. Could you provide more information about this?

Thanks a lot

[tijyojwad]

Hi @guangzhaocs - there's no way to prevent read splitting in dorado. What is your use case for not wanting them to be split?

You could roughly stitch them back together using the pi:z and sp:i tags. the pi tag mentions the original reads ids, and sp is the split point, so sorting by split point and concatenating will approximately get back the unspilt read.

More details on unspilt reads at https://github.com/nanoporetech/dorado/blob/release-v0.6/documentation/SAM.md#split-read-tags

[guangzhaocs]

Thanks very much. I do not have other questions.

[tijyojwad]

note that by default the adapters in reads are also trimmed off, so each split read will have their adapters removed. to get most of the original read back, you need to run with --no-trim