Implement cluster_otus

[JustGitting]

Hi,

I'm keen to use vsearch in place of usearch. A program I'm trying out is IMTORNADO which depends on USEARCH. I tried to substitute USEARCH with VSEARCH, but encountered the following error (imtornado:issue2) :

vsearch: unrecognized option '-cluster_otus'

Does vsearch offer such functionality or does the function have a different option name in vsearch compared to usearch? Or will the function be offered in future versions of vsearch?

Cheers

[torognes]

Hi @JustGitting !

The cluster_otus command is not implemented in vsearch. It performs a combination of clustering and chimera detection as far as I know. We may include it in a future version, but there are no firm plans to do so.

• Torbjørn

[JustGitting]

Hi @torognes

Thanks for your quick reply.

The USEARCH website lacks details about how cluster_otus works. Did you have any other references that would help clarify the algorithm used?

Regards

[colinbrislawn]

Hello @JustGitting

The usearch website does indeed lack a clear description of the --cluster_otus algorithm. Perhaps these pages will help: http://www.drive5.com/usearch/manual/uparseotu_algo.html http://www.drive5.com/usearch/manual/uparseref_algo.html

Here is my understanding of --cluster_otus; it starts by clustering similar sequences just like --cluster_smallmem, but only allows an unmatched read to become the centroid of a new OTU after it has passed the stringent --uparse_ref command. The exact method is unclear.

It looks like imtornado2 now supports vsearch! Perhaps you could ask them about this command. They must have a workaround! https://github.com/pjeraldo/imtornado2#version-203

[JustGitting]

Thanks @colinbrislawn, but it would appear to be only partial support at this stage.

[torognes]

Some details are available in the UPARSE paper, mostly in the online methods:

http://www.nature.com/nmeth/journal/v10/n10/full/nmeth.2604.html