-
StrAuto
StrAuto
StrAuto automates and parallelizes Structure analysis. From a single input file it generates the Structure parameter and command files, runs the full ensemble of K values × replicate runs, and chains the results into the Evanno ΔK analysis with StructureHarvester and CLUMPP.
License
Free to use and open source under GNU GPLv3.
Available
- Roihu: 1.0 (module
strauto), via thebio-appsmodule.
Loading the strauto module also loads the tools it drives — structure,
structureharvester and clumpp — so you do not need to load them separately.
Usage
StrAuto is part of the bio-apps collection on Roihu. Load the bio-apps module tree and then the StrAuto module:
On Roihu, StrAuto is run through the strauto-roihu launcher, which submits the
Structure ensemble as a native Slurm array job.
Move to the scratch directory of your project (any subdirectory inside your scratch area will do) and create a new empty working directory:
StrAuto uses two input files: a parameter file that must be named input.py, and a
genotype data file whose name is defined in input.py and ends with .str or
.ustr. Template files are provided by the module through the $STRAUTO_TEMPLATES
environment variable, and can be copied to your working directory:
Edit input.py to match your analysis. Note that parallelization is handled by the
Slurm array job, so you should not set the parallel parameter to True in
input.py.
When the input file is ready, launch the analysis:
This generates the Structure parameter and command files and submits the full K × replicate ensemble as a Slurm array job.
You can check on the analysis by running the same command again in the same working directory:
This reports the status of the Structure jobs and, once all the Structure runs have finished, performs the post-processing of the results (StructureHarvester Evanno ΔK and CLUMPP).
For long-running analyses, use strauto-roihu-longrun, which submits the job to the
longrun partition: