PASTEC-singularity

Description

Singularity container for the transposable elements classification tool PASTEC(from the package REPET).

Dependancies

• singularity (tested with v.3.9.5)

Installation

Development installation

• Clone this repo:

git clone https://github.com/TommasoBarberis/PASTEC-singularity.git

• Move to the PASTEC directory:

cd PASTEC-singularity

• To build the .sif image from the definition file:

sudo singularity build pastec_latest.sif PASTEC.def

Only user installation

• Pull directly the .sif image from the sylabs.io registry:

singularity pull --arch amd64 library://tommasobarberis98/pastec/pastec:latest

Utilisation

Store your project directory path in a variable:

PROJECT_DIR=path/to/project/dir

This directory has to contain:

• the fasta file with consensi sequences:

  From the PASTEC documentation, each sequence must have 60 bp (or less) by line, so if you have a "one-line" FASTA file, you can transform it with the following command:

  sed '/^>/!s/.\{60\}/&\n/g' oneline.fasta > multiline.fasta
  

  About the sequence headers, it is highly advised to write them like this : ">XX_i" with XX standing for letters and i standing for numbers.
  So, if you another type of header, you can try:

  sed 's/>.\+$/>TE/g' old_file.fa | awk '/^>/{$0=$0"_"(++i)}1' > new_file.fa
  

• the PASTEClassifier.cfg file (a template is provided with the github repo at the root folder) to update some parameters such as the path to a known database of transposable elements (see more here: PASTEClassfier-tuto). Options that you are suggested to update:

  • project_name: whatever you want.
  • TE_nucl_bank: /mnt/nucl_bank.fa such as repbase20.05_ntSeq_cleaned_TE.fa (from RepBase20.05_REPET.embl require subscription to girinst), but you are free to use any other database. Then you can set:
    • TE_BLRtx: yes ➡️ homology with known TEs using tblastx.
    • TE_BLRn: yes ➡️ detection of helitron extremities.
  • TE_prot_bank: /mnt/prot_bank.fa such as repbase20.05_aaSeq_cleaned_TE.fa (from RepBase20.05_REPET.embl require subscription to girinst), but you are free to use any other database. Then you can set:
    • TE_BLRx: yes ➡️ homology with known TEs using blastx.
  • HG_nucl_bank: /mnt/host_genes.fa for the cDNA database of the host genes. Then you can set:
    • HG_BLRn: yes ➡️ homology with host genes.
  • rDNA_bank: /mnt/rdna.fa for the rDNA database. Then you can set:
    • rDNA_BLRn: yes ➡️ homology with rDNA.
  • TE_HMM_profiles: is already set to the ProfilesBankForREPET_Pfam32.0.hmm database, but your are free to use an another database.
  • Adjustable parameters in [classif_consensus] section (see PASTEClassfier-tuto).

• the various databases that you can define in the PASTEClassifier.cfg file.

Container initialization

The PASTEC program use the MySQL database that need to be run as a server, for that you need to initialize the container with this service by creating a singularity instance:

bash path/to/init.sh -s pastec_latest.sif -d $PROJECT_DIR

Interactive mode

• Start a shell through the container:

cd $PROJECT_DIR
singularity shell instance://pastec

• Once in the container, run:

python2.7 /opt/PASTEC_linux-x64-2.0/bin/PASTEClassifier.py -i /mnt/consensi.fasta -C /mnt/PASTEClassifier.cfg

NOTE: the /mnt refer to your PROJECT_DIR, so you have to conserve it and you have only to append the file names.

Batch mode

Create a file pastec_cmd.sh (a template is provided in the test folder) that will contain the PASTEC command in your project directory ($PROJECT_DIR):

#! /bin/bash 

cd /mnt
python2.7 /opt/PASTEC_linux-x64-2.0/bin/PASTEClassifier.py -i /mnt/consensi.fasta -C /mnt/PASTEClassifier.cfg

Then you can run PASTEC using the container instance as follow:

cd $PROJECT_DIR
singularity exec instance://pastec /mnt/pastec_cmd.sh

NOTE: the /mnt refer to your PROJECT_DIR, so you have to conserve it and you have only to append the file names.

Other PASTEC options

• classification rules file name (e.g. PASTEClassifierRules.yaml) [optional]

-D DECISIONRULESFILENAME, --decisionRules=DECISIONRULESFILENAME

• step (0/1/2): default: 0 for all steps, step 1 for detect features, step 2 for classification run tool in parallel

-S STEP, --step=STEP

• clean temporary files [optional] [default: False]

-c, --clean

• verbosity [optional] [default: 3, from 1 to 4]

-v VERBOSITY, --verbosity=VERBOSITY