Accessing software and packages
By default, modules are not loaded onto our individual work space. If we want to use a module or package, we will need to “load” it. Note, a module is a software such as R or python, and a package is run within a module (e.g. dplyr in R or pandas in python).
Before we start using modules and packages, we should understand the reasoning why they are not auto loaded. The two biggest factors are incompatibilities and versioning.
Module incompatibility is a major headache for programmers.
Sometimes the presence (or absence)
of a package will break others that depend on it.
Two of the most famous examples are Python 2 and python 3 and C compiler versions.
Python 3 famously provides a python command
that conflicts with that provided by Python 2.
Software compiled against a newer version of the C libraries
and then used when they are not present will result in a nasty
'GLIBCXX_3.4.20' not found error, for instance.
Module versioning is another common issue. A team might depend on a certain package version for their research project - if the module version was to automatically change (for instance, if it was updated), it might affect their results. Having access to multiple software versions allow a set of researchers to take software version out of the equation if results are weird.
What modules can we access on the HPC
To see available modules on the HPC, use module avail command. Your terminal needs to be logged into the HPC.
[s1234567@gc-prd-hpclogin1 ~]$ module avail
--------------------------------- /usr/share/Modules/modulefiles ---------------------------------
dot module-git module-info modules null use.own
---------------------------------------- /etc/modulefiles ----------------------------------------
mpi/mpich-3.2-x86_64
------------------------------------------ /sw/Modules -------------------------------------------
anaconda2/2019.07py2 library/hdf5/5-1.10.5-intel
anaconda3/2019.07py3 library/lapack/3.8.0
ansys/v195 library/loki/0.1.7
bioinformatics/blast+/2.9.0 library/netcdf/netcdf-c/4.7.3
bioinformatics/bowtie2/2.3.5.1 library/netcdf/netcdf-c/4.7.3-intel
bioinformatics/cd-hit/4.8.1 library/netcdf/netcdf-cxx/4.3.1
bioinformatics/celSEQpipeline/1.0 library/netcdf/netcdf-cxx/4.3.1-intel
bioinformatics/fastqc/0.11.8 library/netcdf/netcdf-fortran/4.5.2
bioinformatics/htslib/1.9 library/netcdf/netcdf-fortran/4.5.2-intel
bioinformatics/orthofinder/2.3.7 library/openblas/0.3.6
bioinformatics/samtools/1.9 library/proj/6.2.1
bioinformatics/signalp/4.1 library/scalapack/2.0.2
bioinformatics/tmhmm/2.0 library/tcl/8.6.9
bioinformatics/transdecoder/5.5.0 library/tk/8.6.9
bioinformatics/trinity/2.8.6 library/vtk/8.2.0
cmake/3.15.5 library/xz/5.2.4
cmake/3.8.2 matlab/2018b
delft3d/6244intel matlab/2019b
extras/jags/4.3.0 misc/aem3d/1.0.1
gaussian/g16 misc/aem3d/aem3d
gcc/8.3.1 misc/automake/1.15
gcc/old/9.2.0 misc/baronsolver/19.7.13
gcc/old/9.2.0OLD misc/cplex/cplex
gromacs/2019.3 misc/libtool/2.4.6
gromacs/2019.4 misc/metashape/1.5.2
gromos/md++141 misc/metashape/1.5.3pro
intel/ics2013 misc/metis/5.1.0
intel/intelmpi misc/qglviewer/2.6.3
intel/intelParallelStudio2019 misc/qglviewer/2.7.1
java/jdk11.0.1 misc/simbody/361
lang/golang/1.21.0 misc/yade/2019.01a
lang/golangci-lint/1.21.0 misc/yasara/yasara
legacy/gcc/4.9.3 mpi/mpich/3.3.2-gnu
library/boost/1.59.0py3 mpi/mpich/3.3.2-intel
library/boost/1.70.0py2 mpi/openmpi/2.1.6
library/boost/1.70.0py3 mpi/openmpi/4.0.2
library/cgal/4.13.1 python/2.7.17
library/eigen/3.3.7 python/3.7.4
library/fftw/3.3.8 qt/5.12.3
library/fftw/3.3.8mpi quantum-espresso/6.4.1
library/gdal/3.0.2 R/3.6.1
library/geos/3.8.0 singularity/3.2.0
library/gts/0.7.6 vasp/vasp541-gnu-openmpi
library/hdf4/4.2.13 virtualgl/2.6.3
library/hdf5/5-1.10.5
If the module you require is not avaliable on the HPC then fill in this form and in “Problem Description” write the software/package you would like along with the version number.
Loading and unloading software
Intially, the statistical program R is not loaded.
[s1234567@gc-prd-hpclogin1 ~]$ which R
no R in (/opt/clmgr/sbin:/opt/clmgr/bin:/opt/sgi/sbin:/opt/sgi/bin:/usr/local/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/opt/c3/bin:/opt/pbs/bin:/sbin:/bin:/export/home/sNumber/bin)
Load R into your environment using the module load command. If you remember from the scheduler tutorial, we can add module load into the scheduler script.
[s1234567@gc-prd-hpclogin1 ~]$ module load R
[s1234567@gc-prd-hpclogin1 ~]$ R --version
R version 3.6.1 (2019-07-05) -- "Action of the Toes"
Copyright (C) 2019 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under the terms of the
GNU General Public License versions 2 or 3.
For more information about these matters see
https://www.gnu.org/licenses/.
What if we want a specific version of a program? There are 2 versions of python avaliable though, 2.7.17 and 3.7.4. Lets load python 3.7.4.
[s1234567@gc-prd-hpclogin1 ~]$ module load python/3.7.4
[s1234567@gc-prd-hpclogin1 ~]$ python --version
Python 3.7.4
Success!!!
module load will add software to your $PATH.
To see which modules are loaded we can run the module list command. Note this will not list packages within modules (i.e. it won’t show dplyr in R or pandas in python)
[s1234567@gc-prd-hpclogin1 ~]$ module list
Currently Loaded Modulefiles:
1) python/3.7.4
[s1234567@gc-prd-hpclogin1 ~]$ module load R
[s1234567@gc-prd-hpclogin1 ~]$ module list
Currently Loaded Modulefiles:
1) R/3.6.1 2) python/3.7.4
So in this case, the software R and python were loaded into your HPC workspace.
Let’s try unloading the R software.
[s1234567@gc-prd-hpclogin1 ~]$ module unload R
[s1234567@gc-prd-hpclogin1 ~]$ module list
Currently Loaded Modulefiles:
1) python/3.7.4
So using module unload “un-loads” a module along with it’s dependencies.
If we wanted to unload everything at once, we could run module purge (unloads everything).
[s1234567@gc-prd-hpclogin1 ~]$ module purge
No Modulefiles Currently Loaded.
Note that module purge is informative.
It lets us know that all but a default set of packages have been unloaded
(and how to actually unload these if we truly so desired).
Software versioning
So far, we’ve learned how to load and unload software packages. This is very useful. However, we have not yet addressed the issue of software versioning. At some point, you will run into issues where only one particular version of some software will be suitable. Perhaps a key bugfix only happened in a certain version, or version X broke compatibility with a file format you use. In either of these example cases, it helps to be very specific about what software is loaded.
Let’s examine the output of module avail more closely.
[s1234567@gc-prd-hpclogin1 ~]$ module avail
--------------------------------- /usr/share/Modules/modulefiles ---------------------------------
dot module-git module-info modules null use.own
---------------------------------------- /etc/modulefiles ----------------------------------------
mpi/mpich-3.2-x86_64
------------------------------------------ /sw/Modules -------------------------------------------
anaconda2/2019.07py2 library/hdf5/5-1.10.5-intel
anaconda3/2019.07py3 library/lapack/3.8.0
ansys/v195 library/loki/0.1.7
bioinformatics/blast+/2.9.0 library/netcdf/netcdf-c/4.7.3
bioinformatics/bowtie2/2.3.5.1 library/netcdf/netcdf-c/4.7.3-intel
bioinformatics/cd-hit/4.8.1 library/netcdf/netcdf-cxx/4.3.1
bioinformatics/celSEQpipeline/1.0 library/netcdf/netcdf-cxx/4.3.1-intel
bioinformatics/fastqc/0.11.8 library/netcdf/netcdf-fortran/4.5.2
bioinformatics/htslib/1.9 library/netcdf/netcdf-fortran/4.5.2-intel
bioinformatics/orthofinder/2.3.7 library/openblas/0.3.6
bioinformatics/samtools/1.9 library/proj/6.2.1
bioinformatics/signalp/4.1 library/scalapack/2.0.2
bioinformatics/tmhmm/2.0 library/tcl/8.6.9
bioinformatics/transdecoder/5.5.0 library/tk/8.6.9
bioinformatics/trinity/2.8.6 library/vtk/8.2.0
cmake/3.15.5 library/xz/5.2.4
cmake/3.8.2 matlab/2018b
delft3d/6244intel matlab/2019b
extras/jags/4.3.0 misc/aem3d/1.0.1
gaussian/g16 misc/aem3d/aem3d
gcc/8.3.1 misc/automake/1.15
gcc/old/9.2.0 misc/baronsolver/19.7.13
gcc/old/9.2.0OLD misc/cplex/cplex
gromacs/2019.3 misc/libtool/2.4.6
gromacs/2019.4 misc/metashape/1.5.2
gromos/md++141 misc/metashape/1.5.3pro
intel/ics2013 misc/metis/5.1.0
intel/intelmpi misc/qglviewer/2.6.3
intel/intelParallelStudio2019 misc/qglviewer/2.7.1
java/jdk11.0.1 misc/simbody/361
lang/golang/1.21.0 misc/yade/2019.01a
lang/golangci-lint/1.21.0 misc/yasara/yasara
legacy/gcc/4.9.3 mpi/mpich/3.3.2-gnu
library/boost/1.59.0py3 mpi/mpich/3.3.2-intel
library/boost/1.70.0py2 mpi/openmpi/2.1.6
library/boost/1.70.0py3 mpi/openmpi/4.0.2
library/cgal/4.13.1 python/2.7.17
library/eigen/3.3.7 python/3.7.4
library/fftw/3.3.8 qt/5.12.3
library/fftw/3.3.8mpi quantum-espresso/6.4.1
library/gdal/3.0.2 R/3.6.1
library/geos/3.8.0 singularity/3.2.0
library/gts/0.7.6 vasp/vasp541-gnu-openmpi
library/hdf4/4.2.13 virtualgl/2.6.3
library/hdf5/5-1.10.5
We can also see if there is a specific module available
[s1234567@gc-prd-hpclogin1 ~]$ module avail R
--------------------------------- /sw/Modules ----------------------------------
R/3.6.1 R/4.0.3
----------------------------- /sw/centos7/Modules ------------------------------
R/3.5.1
Let’s have a closer look at the gcc module.
GCC is a widely used C/C++/Fortran compiler.
Alot of software is dependent on the GCC version,
and compile or run issues can occur if the wrong version is loaded.
There are three different versions avaliable on the HPC: gcc/8.3.1, gcc/old/9.2.0 and gcc/old/9.2.0OLD.
How do we load each copy and which copy is the default?
[s1234567@gc-prd-hpclogin1 ~]$ module load gcc
gcc --version
gcc (GCC) 9.2.0
Copyright (C) 2019 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
Three things happened here:
- 1) the default copy of GCC was loaded (version 9.2.0),
- 2) the Intel compilers (which conflict with GCC) were unloaded,
- 3) and software that is dependent on compiler (OpenMPI) was reloaded.
The
modulesystem turned what might be a super-complex operation into a single command.
So how do we load the non-default copy of a software package?
In this case, the only change we need to make is be more specific about the module we are loading.
There are two GCC modules: gcc/8.3.1 and gcc/old/9.2.0OLD.
To load a non-default module, the only change we need to make to our module load command
is to add in the required version number after the /.
Load a specific version of GCC
Load GCC version 7.3.0
Solution
[s1234567@gc-prd-hpclogin1 ~]$ module unload gcc[s1234567@gc-prd-hpclogin1 ~]$ module load gcc/7.3.0 [s1234567@gc-prd-hpclogin1 ~]$ gcc --versiongcc (GCC) 7.3.0 Copyright (C) 2017 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
Identify the installed packages within a specific modules
First you need to load and initiate R:
[s1234567@gc-prd-hpclogin1 ~]$ module load R
[s1234567@gc-prd-hpclogin1 ~]$ R
R version 3.6.1 (2019-07-05) -- "Action of the Toes"
Copyright (C) 2019 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.
Natural language support but running in an English locale
R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.
Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.
Then you can see which packages are avaliable. Below is an example of using R to create a list of the packages.
> ip <- as.data.frame(installed.packages()[, c(1, 3:4)])
> rownames(ip) <- NULL
> ip <- ip[is.na(ip$Priority), 1:2, drop=FALSE]
> print(ip, row.names=FALSE)
Package Version
abind 1.4-5
acepack 1.4.1
anchors 3.0-8
askpass 1.1
assertthat 0.2.1
backports 1.1.5
base64enc 0.1-3
BH 1.69.0-1
brew 1.0-6
broom 0.5.2
callr 3.3.2
cellranger 1.1.0
checkmate 2.0.0
classInt 0.4-2
cli 1.1.0
clipr 0.7.0
clisymbols 1.2.0
... ...
Then quit R to get back to the terminal.
> q()
Save workspace image? [y/n/c]: n
[s1234567@gc-prd-hpclogin1 ~]$
Notice that once you run R the adress changes from [s1234567@gc-prd-hpclogin1 ~]$ to >, signifying that we are no longer in our home directory on the HPC, but are instead in an instance of R on the HPC.
For python:
[s1234567@gc-prd-hpclogin1 ~]$ module load python/3.7.4
[s1234567@gc-prd-hpclogin1 ~]$ python -c 'help("modules")'
Note that we do not run python to identify the packages.
Please wait a moment while I gather a list of all available modules...
/sw/python/3.7.4/lib/python3.7/site-packages/pandas/compat/__init__.py:117: UserWarning: Could not import the lzma module. Your installed Python is incomplete. Attempting to use lzma compression will result in a RuntimeError.
warnings.warn(msg)
Cython asynchat imghdr rlcompleter
HTSeq asyncio imp runpy
__future__ asyncore importlib sched
_abc atexit importlib_metadata secrets
_ast attr inspect select
_asyncio audioop io selectors
_bisect backports ipaddress setuptools
_blake2 base64 ipython_genutils shelve
_bootlocale bdb itertools shlex
_bz2 binascii json shutil
_codecs binhex jsonschema signal
_codecs_cn bisect jupyter site
... ... ... ...
If the module you require is not avaliable on the HPC then fill in this form and in “Problem Description” write the software/package you would like along with the version number.
Anaconda environments
Anaconda is a distribution of python and R designed specifically for scientific computing. Anaconda contains the conda command, a package and environment manager that aims to manage dependencies and isolate projects. It is really cool, and really useful, for more indepth information have a look here
There are multiple prebuilt conda environments avaliable on Griffith’s HPC that are suited to bioinformatics.
There is a python 2.x and 3.x installation of Anaconda avaliable
[s1234567@gc-prd-hpclogin1 ~]$ module avail
anaconda2/2019.07py2
anaconda3/2019.07py3
Load the module of interest and see what environments are avaliable
[s1234567@gc-prd-hpclogin1 ~]$ module load anaconda3/2019.07py3
[s1234567@gc-prd-hpclogin1 ~]$ conda info --envs
# conda environments:
#
base * /sw/anaconda3/2019.07
3point6 /sw/anaconda3/2019.07/envs/3point6
R /sw/anaconda3/2019.07/envs/R
bioinformatics /sw/anaconda3/2019.07/envs/bioinformatics
bioinformatics2 /sw/anaconda3/2019.07/envs/bioinformatics2
cellprofiler /sw/anaconda3/2019.07/envs/cellprofiler
roaryENV /sw/anaconda3/2019.07/envs/roaryENV
toxify_env /sw/anaconda3/2019.07/envs/toxify_env
trinityENV /sw/anaconda3/2019.07/envs/trinityENV
Load the environment of interest and check the packages and versions avaliable
[s1234567@gc-prd-hpclogin1 ~]$ source activate bioinformatics
(bioinformatics) [s1234567@gc-prd-hpclogin1 ~]$ conda list
# packages in environment at /sw/anaconda3/2019.07/envs/bioinformatics:
#
# Name Version Build Channel
_libgcc_mutex 0.1 main
a5pipeline 20150522 1 vdbwrair
aragorn 1.2.36 0 biocore
barrnap 0.9 3 bioconda
bcftools 1.9 h68d8f2e_9 bioconda
bedtools 2.29.2 hc088bd4_0 bioconda
blast 2.9.0 pl526h3066fca_4 bioconda
blast-plus 2.2.31 0 biocore
bowtie 1.0.0 1 bioconda
bowtie2 2.2.5 0 bioconda/label/broken
bwa 0.7.17 hed695b0_7 bioconda
bzip2 1.0.8 h7b6447c_0
ca-certificates 2020.1.1 0
cairo 1.16.0 hfb77d84_1002 conda-forge
cd-hit 4.8.1 h8b12597_3 bioconda
certifi 2019.11.28 py38_0
collectl 4.0.4 2 bioconda
configparser 4.0.2 pypi_0 pypi
curl 7.68.0 hbc83047_0
... ... ... ...
To deactivate the environment
[s1234567@gc-prd-hpclogin1 ~]$ conda deactivate
Build your own anaconda environment
You can create your own conda environment in your home directory but please make sure that it is not already avaliable. Creating a conda environment requires internet access.
# To get internet access from the cluster, run this command:
[s1234567@gc-prd-hpclogin1 ~]$ source /usr/local/bin/s3proxy.sh
# Load the require anaconda environment, i.e. 2 or 3
[s1234567@gc-prd-hpclogin1 ~]$ module load anaconda3
# create a new conda environment using the create command. Use the -n flag to give it a name, and tell it the version of python and anaconda to create the environment with
[s1234567@gc-prd-hpclogin1 ~]$ conda create -n NameOfCondaEnv python=X.X anacondaX
# There will now be a new conda environment with that name
[s1234567@gc-prd-hpclogin1 ~]$ conda info --envs
# activate the environment
[s1234567@gc-prd-hpclogin1 ~]$ source activate NameOfCondaEnv
# Install the packages you would like to use
[s1234567@gc-prd-hpclogin1 ~]$ conda install -n NameOfCondaEnv [package]
# deactivate the environment (older versions will use source deactivate)
[s1234567@gc-prd-hpclogin1 ~]$ conda deactivate
# once you have finished with it and will not be using it again you can delete it
[s1234567@gc-prd-hpclogin1 ~]$ conda env remove -n NameOfCondaEnv