gravatar for malteherold

2 hours ago by

Hello,

In trying to do typing of bacterial strains from metagenomic samples I recently came across this tool, strainEST:
github.com/compmetagen/strainest
www.nature.com/articles/s41467-017-02209-5

It uses mapping against an alignment of representative strains and SNV profiles to predict relative abundances of individual strains in metagenomic samples.
As the species I am interested in does not have a pre-built database, the first step was to construct this.
After reducing around 3500 reference strains to 650 clusters or representatives, which is quite a lot compared to what is used in the paper, I am now stuck in generating the alignments and am now also wondering if with this number the following steps will be feasible at all.

Stuck at this point (at 92% for hours, so close...):


[email protected]:/strainest/representative_strains# strainest mapgenomes *fasta  SR.fa MR.fasta
  [#################################---]   92%  0d 00:10:04

Does anyone have experience with this tool and knows what kind of resources are required for database construction and running the tool with a larger amount of representative strains?
Also are there any comparable tools that you can recommend for the same purpose (identifying and tracking a large number of bacterial strains in few metagenomic samples)?

link

modified 4 minutes ago

written
2 hours ago
by

malteherold20



Source link