BioMonth - CSC supercomputing and data management for bioscientists

Installing applications and libraries for Java, Perl, Python and R

Java

Java applications do not typically require special installation. You just need to download the .jar file.

If you get an error message about java version, try loading a suitable java module. You can check the available modules with command:

module spider java

Despite their name, modules named biojava are just normal java installations, and can be used with any software, not just bio applications.

Some java applications fail to run on the login nodes. Try running them using sinteractive instead.

Naturally no compute heavy tasks should be run on the login nodes.

Perl

To run Perl applications it usually best to first load a perl module.

To check available versions, use:

module spider perl

To install and use your own perl packages, it usually enough to download them to a suitable location and make sure they are included in perl @INC

There are three main ways to do it:

1. Inlude command line option -I (capital i) with the path on the command line:

   perl -I /projappl/project_12345/myperl ./my_app.pl
   

2. Include the path in $PERL5LIB environment variable.

   export PERL5LIB=/projappl/project_12345/myperl:${PERL5LIB}
   

3. Iclude the path in the perl code with use lib

    use lib '/projappl/project_12345/myperl';
    use My::Module;
   

   Bioperl

   If your code requires bioperl, we have a bioperl installation available.

See our biperl documentation for details.

Python

To run Python applications, first load suitable Python module. You can check available modules with e.g.:

module spider python

• python-env is a general purpose Python installation
• python-data includes commonly used packages for data analysis and machine learning

To install simple packages it is usually enough to use:

pip install --user package_to_install

Remember to include --user. By default it tries to install to Python install location, and this will not work.

For more complex installations it is preferable to create a virtual environment.

See instructions in our python-data documentation. The same instructions work also for the python-env module.

Biopython

For applications requiring Biopython we have two options:

First option using biopythontools module:

module load biokit
module load biopythontools

With this option use pip install --user as above.

Second option activates a virtual environment with biopython (substitute your project name):

export PROJAPPL=/projappl/project_12345
module load bioconda
biopython-env

See our Biopython documentation for more details.

R

Puhti has several R version available as modules. You should start by checking if one of these is suitable for you needs.

To check available versions, use:

module spider r-env

It is also possible to install your own R packages.

See the r-env-singularity documentation for instructions.

This site is open source. Improve this page.