Whitelist subreads instead of ZMWs?

[Aizon]

Hello -

I was wondering whether it is possible, in the whitelist.txt to use the format:

<movie>/<zmw>/<coord>_<coord>

rather than the typical:

<movie>/<zmw>

If not, what other ways are there to eliminate specific subreads, rather than the entire zmw?

Thanks,
A

[rhall]

The short answer is no.

Whitelisting has been primarily used for contamination removal so per molecule has always been the expected use case.

What is the intention of removing one subread and not another from the same molecule? Could you achieve the same outcome using whitelist and minimum subread length?

[Aizon]

Hello rhall and thanks for your reply.

To answer as shortly as possible:
Due to some assembly issues, I want to select for assembly only those subreads that map to a reference genome for a minimum ratio of their length.

If a 12 Kb subread only gets 3 Kb mapping to the reference, filter out.
If a 3.5 Kb subread from the same hole gets 3 Kb mapping to the reference, keep it.

I was thinking/wondering, that it might be an overkill to remove the whole ZMW.

[rhall]

So long as you are not really coverage limited, filtering by ZMW is probably the most straight forward.
Otherwise you could use an assembler that allows fasta input, https://github.com/PacificBiosciences/FALCON or http://wgs-assembler.sourceforge.net...index.php/PBcR

[Aizon]

I see. Will consider my options!

Thanks for the fast replies.

Cheers!