Cluster.split issues - splitmethod not valid

[MDelavy]

Hi everyone,

I’m trying to perform cluster.split on my sequences (3M unique sequences) as described in the MiSeq SOP (but at the genus level).

cluster.split(fasta=16S_data/stability.trim.contigs.good.unique.good.filter.unique.precluster.pick.pick.fasta,count=16S_data/stability.trim.contigs.good.unique.good.filter.unique.precluster.denovo.vsearch.pick.pick.count_table, taxonomy=16S_data/stability.trim.contigs.good.unique.good.filter.unique.precluster.pick.nr_v138_1.wang.pick.taxonomy,cluster=f,taxlevel=6,cutoff=0.03,processors=6)

And I’m facing the following issues:

First, I got an error message stating : “splitmethod is not a valid parameter, ignoring”. I didn’t find any records of this issue anywhere

Then, despite the fact I specified the number of processors I would like to use, it doesn’t seems like there are any parallel work. Only a single counttemp and dist file is produced at the same time. Is it normal ?

I used this command in another project without facing these issues so I’m worried something is wrong here. Could you help me ?

Thanks a lot,
Margot

(By the way, I’m using Mothur 1.46.1)

[westcott]

Hi Margot,

In our latest release (1.46.1) we made some improvements to the cluster.split command. We reduced timing of cluster.split through improvements in the splitting process and removed split by distance option from cluster.split. The splitmethod parameter was removed because there is only one splitting method, taxonomy.

The cluster.split command uses one processor for the splitting of the sequences into taxonomic groups, multiple processors when calculating the distance files for the individual groups. The clustering process uses multiple processors, but the assembling of the completed list file and sens.spec results use one processor.

Kindly,
Sarah Westcott