.. _`tpm-immunedeconv (✓ production)`:

TPM-IMMUNEDECONV (✓ PRODUCTION)
===============================

Version: 11-07-2019
Tags: RNA-Seq / TPM / Deconvolution / immunedeconv

Immunedeconv is based on many methods: EPIC, Cibersort, Cibersort.ABS, quantiseq, XCell, and MCP-Counter. Choose which tool to run in the configuration files.

You may find more information at

   * https://github.com/icbi-lab/immunedeconv/blob/master/README.md
   * https://github.com/tdayris/tpm-immunedecov


.. role:: bash(code)
  :language: bash

Pipeline dependencies
---------------------

This pipeline requires the following packages to be run. Any other additional requirements are being installed dynamically.

Conda:
  
  * conda-forge::python=3.8.2
  
  * conda-forge::pytest=5.4.1
  
  * conda-forge::datrie=0.8.2
  
  * conda-forge::git=2.26.0
  
  * conda-forge::jinja2=2.11.1
  
  * conda-forge::pygraphviz=1.5
  
  * conda-forge::flask=1.1.1
  
  * conda-forge::pandas=1.0.3
  
  * conda-forge::zlib=1.2.11
  
  * conda-forge::openssl=1.1.1e
  
  * conda-forge::networkx=2.4
  
  * bioconda::snakemake=5.16.0
  


Additionally, the following prerequisites are non-optional:


* Conda

* Genome sequence and annotation




Input files
-----------

Please find below the list of required input files:


* TSV-formatted **TPM** counts matrix



Output files
------------

Please find below the list of expected output files:


* TSV formatted deconvolution results estimation

* PNG images representing the population

* Complete quantification report embedding material and methods




Notes
-----

This pipeline takes the cold storage into account. No need to copy your data in advance.

In order to install, use "conda" to install required environment, and "git" to clone the git repository.





Installation
------------

While installing the workflow, you may run the following commands (order matters):

.. list-table::
  :widths: 10 80
  :header-rows: 1
  :align: left

  * - Case
    - Command line
  * - git
    - .. code-block:: bash 

        # This command clones the github repository

        if [ ! -d "${TPM_IMMUNEDECONV_DIR:?}" ]; then git clone https://github.com/tdayris/tpm-immunedeconv.git "${TPM_IMMUNEDECONV_DIR:?}"; fi
  * - conda
    - .. code-block:: bash 

        # This command creates the conda virtual environment. It requires an

        # access to the git repository (see above).

        conda env create --force --file "${STRONGR_DIR:?}/workflows/genomic-expression/tpm-immunedeconv/environment.yaml"



Testing
-------

In order to test the pipeline, you may try the following commands:

.. list-table::
  :widths: 10 80
  :header-rows: 1
  :align: left

  * - Case
    - Command line
  * - quick-test
    - .. code-block:: bash 

        cd "${TPM_IMMUNEDECONV_DIR:?}"

        make conda tests

        make all-unit-tests

        make test-conda-report.html

        make clean




Preparation
-----------

In order to prepare a run, you may try the following commands:

.. list-table::
  :widths: 10 80
  :header-rows: 1
  :align: left

  * - Case
    - Command line
  * - activate
    - .. code-block:: bash 

        # This command activates the conda environment available after the

        # installation process.

        conda activate tpm-immunedeconv || source activate tpm-immunedeconv
  * - gustaveroussy-references
    - .. code-block:: bash 

        # This points mount point configurations for Gustave Roussy's flamingo

        COLD_STORAGE=(/mnt/isilon /mnt/archivage)



Execution
---------

In order to execute the pipeline, you may run the following commands:

.. list-table::
  :widths: 10 80
  :header-rows: 1
  :align: left

  * - Case
    - Command line(s)
  * - local
    - .. code-block:: bash 

        source activate tpm-immunedeconv || conda activate tpm-immunedeconv

        snakemake -s "${TPM_IMMUNEDECONV_DIR:?}/Snakefile" -r -p --configfile config.yaml -j 4 --use-conda
  * - torque
    - .. code-block:: bash 

        # While reserving optimal threads and memory requirements,

        # the choice of the queue might not be optimal.

        # See profiles below.

        source activate tpm-immunedeconv || conda activate tpm-immunedeconv

        snakemake -s "${TPM_IMMUNEDECONV_DIR:?}/Snakefile" -r -p --configfile config.yaml -j 100 --cluster "qsub -V -d ${TPM_IMMUNEDECONV_WORKDIR:?} -j oe -l nodes=1:ppn={threads},mem={resources.mem_mb}mb,walltime={resources.time_min}:00" --use-conda
  * - slurm
    - .. code-block:: bash 

        # While reserving optimal threads and memory requirements,

        # the choice of the queue might not be optimal.

        # See profiles below.

        source activate tpm-immunedeconv || conda activate tpm-immunedeconv

        snakemake -s "${TPM_IMMUNEDECONV_DIR:?}/Snakefile" -r -p --configfile config.yaml -j 100 --cluster "sbatch --mem={resources.mem_mb} --time={resources.time_min} --cpus-per-task={threads}" --use-conda
  * - profile
    - .. code-block:: bash 

        {'\# Requires slurm profile installation, see': None}

        source activate tpm-immunedeconv || conda activate tpm-immunedeconv

        snakemake -s "${TPM_IMMUNEDECONV_DIR:?}/Snakefile" --configfile config.yaml --profile slurm
  * - report
    - .. code-block:: bash 

        snakemake -s "${TPM_IMMUNEDECONV_DIR:?}/Snakefile" --configfile config.yaml --report