Assembly of nextera mate pair libraries

[agseq]

As the gel-free Nextera mate-pair method generates a range of mate-pair library insert sizes, how does one handle assembly informatically? If you assemble in an iterative fashion, say 3Kb, 8Kb and 15Kb, should you remove the reads which are identified in each of the assemblies and then feed them back in separately? Any suggestions are appreciated.

[Featherston]

Hi there

Yeah I am at a complete loss of what to do with this data. None of the scaffolders want to use this data. I've tried SSPACE, SOPRA, OPERA (this doesn't exclude joins based on size), MIP and many assemblers with inbuilt scaffolders. None have worked. The only time I get scaffolding to work at all with the gel-free data is when I include paired-reads together with the mate-pair data as an input. But then I find there are expansions of ambiguities and I get crazy total genome sizes...... This is painful because even though you're joining contigs how does one report on any metrics with a genome that is twice the size it should be? I get decent mapping as well so it's not that my reads aren't mapping. Hope you've made some progress! Did you try the iterative approach?

BTW I have had some success with using NextClip for identifying true mate-pairs.
Cheers
J