Build all the containers inside the build_containers
directory:
cd build
./buildall
When that finishes, it will give you a command you should add to your .bashrc
which will allow the shell to find the newly built modules.
If you rerun the buildall
script it won't delete the previous singularity containers but will overwrite the run scripts allowing you to build the containers on a system you have the rights to build containers on, and rsync the results to a system where you don't and rerunning buildall
to handle the changes to the location. If you do want to get rid of the containers, run the cleanall
script.
While the versions are all defined in the buildall file, there's a version_check
script which will check for updates. Run it like this:
./version_check -av
Updates can be built using the specific tool's build
command followed by the new version although you should check that there's a conda version available if it fails. The modulefile won't be written in the event of such a failure though.
Before running the workflow make sure you have modified parameters.config
to set your Working_Directory and Resource_Directory.
Key resources include reference information, including a .fna
, .cds.fasta.gz
, and .gff.gz
for your reference. For human assemblies chm13 and hg38 reference information can be downloaded and formatted using Utility-chm13pull.sh
and Utility-hg38pull.sh
, respectively. Your genome of interest may be available to similarly download in the correct format from the NCBI FTP site. Similarly more updated transcript information in .gff3.gz
format is avaiable from the Ensembl FTP site and human infromation can be downloaded using Utility-ensemblpull.sh
, note that these files may need modification before using in the workflow.