Merge branch 'singularity-podman' into 'master'

Singularity podman

See merge request uniprot-public/unifire!11

README.md

@@ -64,18 +64,24 @@ The only input data that need to be provided are the protein sequence data in mu
 
 ### Usage
 ```
-usage: run_unifire_docker.sh -i <INPUT_FILE> -i <OUTPUT_FOLDER> [-v <VERSION> [-w <WORKING_FOLDER [-c]]]
-        -i: Path to multi-FASTA input file with headers in UniProt FASTA header format, containing at least
-            OX=<taxid>. (Required)
-        -o: Path to output folder. All output files with predictions in TSV format will be available in this
-            folder at the end of the procedure. (Required)
-        -v: Version of the docker image to use, e.g. 2020.2. Available versions are listed under
-            https://gitlab.ebi.ac.uk/uniprot-public/unifire/container_registry. (Optional), DEFAULT: 2020.2
-        -w: Path to an empty working directory.  If this option is not given, then a temporary folder will be
-            created and used to store intermediate files. (Optional)
-        -c: Clean up temporary files. If set, then all temporary files will be cleaned up at the end of the
-            procedure. If no working directory is provided through option -w then the temporary files are cleaned
-            up by default
+usage: ./docker/bin/run_unifire_docker.sh -i <INPUT_FILE> -o <OUTPUT_FOLDER> [-v <VERSION>] [-w <WORKING_FOLDER] [-c]
+          [-s docker|singularity|podman]
+    -i: Path to multi-FASTA input file with headers in UniProt FASTA header format, containing at least
+        OX=<taxid>. (Required)
+    -o: Path to output folder. All output files with predictions in TSV format will be available in this
+        folder at the end of the procedure. (Required)
+    -v: Version of the docker image to use, e.g. 2020.2. Available versions are listed under
+        https://gitlab.ebi.ac.uk/uniprot-public/unifire/container_registry. (Optional), DEFAULT: 2020.4.1
+    -w: Path to an empty working directory.  If this option is not given, then a temporary folder will be
+        created and used to store intermediate files. (Optional)
+    -c: Clean up temporary files. If set, then all temporary files will be cleaned up at the end of the
+        procedure. If no working directory is provided through option -w then the temporary files are cleaned
+        up by default
+    -s: Container software to be used. (Optional), DEFAULT: docker
+        Allowed values:
+        docker: Use Docker to run UniFIRE Docker image
+        singularity: Use Singularity to run UniFIRE Docker image
+        podman: Use Podman to run UniFIRE Docker image
 ```
 
 ### Example
@@ -104,6 +110,51 @@ The application of the UniFIRE Docker image on a complete bacterial proteome wit
  procedure.
 <br/>
 
+### Alternatives to Docker
+For various reasons Docker is not a reasonable solution in a multi-user environment like most HPC clusters. Therefore
+ alternatives like *Singularity* and *Podman* have been tested to run the UniFIRE Docker image. 
+
+#### Singularity
+Instead of Docker, an available Singularity installation can be used to run the UniFIRE docker image. The executable
+"singularity" must be available in the PATH environment variable. The UniFIRE Docker image has been tested successfully
+with Singularity version 3.6.1.
+
+Because the UniFIRE image is big, you may want to use a folder with enough free disk-space
+ (~200 GB) available for temporary and cached files:
+```
+export SINGULARITY_CACHEDIR=/path/to/cache/folder
+export SINGULARITY_TMPDIR=/path/to/tmp/folder
+export SINGULARITY_LOCALCACHEDIR=/path/to/localcache/folder
+```
+
+Run the Docker image with Singularity:
+```
+./docker/bin/run_unifire_docker.sh -i samples/proteins.fasta -o . -s singularity
+```
+    
+ 
+#### Podman
+Instead of Docker, an available Podman installation can be used to run the UniFIRE docker image. The executable
+"podman" must be available in the PATH environment variable. The UniFIRE Docker image has been tested successfully
+with Podman version 2.0.3.
+
+Because the UniFIRE image is big, you may want to use a folder with a larger amount of free disk-space
+ (~200 GB) available for temporary and cached files:
+```
+export TMPDIR=/path/to/tmp/folder
+```
+Run the Docker image with Podman:
+```
+./docker/bin/run_unifire_docker.sh -i samples/proteins.fasta -o . -s podman
+```
+
+For both cases, Singularity and Podman, the resulting output filder will be located in ${run_folder} with the filenames
+```
+predictions_unirule.out
+predictions_unirule-pirsr.out
+predictions_arba.out
+```
+
 ## 2. Run UniFIRE after building it from its source code
 
 ### Prerequisites

docker/Dockerfile

@@ -26,7 +26,7 @@ RUN pip install --upgrade ete3
 COPY scripts /opt/scripts/bin
 RUN chmod 775 /opt/scripts/bin/*.sh
 
-RUN /opt/scripts/bin/update-taxonomy-cache.py
+RUN /opt/scripts/bin/update-taxonomy-cache.sh
 RUN /opt/scripts/bin/download-interproscan.sh
 RUN /opt/scripts/bin/download-unifire.sh
 

docker/bin/run_unifire_docker.sh

@@ -23,11 +23,13 @@ infile=""
 outdir=""
 workdir=""
 cleanworkdir=0
+container_software="docker"
 docker_version="2020.4.1"
 predictionfiles="predictions_unirule.out predictions_arba.out predictions_unirule-pirsr.out"
 
 function usage() {
-    echo "usage: $0 -i <INPUT_FILE> -o <OUTPUT_FOLDER> [-v <VERSION> [-w <WORKING_FOLDER [-c]]]"
+    echo "usage: $0 -i <INPUT_FILE> -o <OUTPUT_FOLDER> [-v <VERSION>] [-w <WORKING_FOLDER] [-c]"
+    echo "          [-s docker|singularity|podman]"
     echo "    -i: Path to multi-FASTA input file with headers in UniProt FASTA header format, containing at least"
     echo "        OX=<taxid>. (Required)"
     echo "    -o: Path to output folder. All output files with predictions in TSV format will be available in this"
@@ -39,10 +41,15 @@ function usage() {
     echo "    -c: Clean up temporary files. If set, then all temporary files will be cleaned up at the end of the"
     echo "        procedure. If no working directory is provided through option -w then the temporary files are cleaned"
     echo "        up by default"
+    echo "    -s: Container software to be used. (Optional), DEFAULT: docker"
+    echo "        Allowed values:"
+    echo "        docker: Use Docker to run UniFIRE Docker image"
+    echo "        singularity: Use Singularity to run UniFIRE Docker image"
+    echo "        podman: Use Podman to run UniFIRE Docker image"
     exit 1
 }
 
-while getopts "i:o:w:c:v:" optionName
+while getopts "i:o:w:c:v:s:" optionName
 do
   case "${optionName}" in
     i) infile=${OPTARG};;
@@ -50,9 +57,26 @@ do
     w) workdir=${OPTARG};;
     v) docker_version=${OPTARG};;
     c) cleanworkdir=1;;
+    s) container_software=${OPTARG};;
   esac
 done
 
+if [ ${container_software} != "docker" ] && [ ${container_software} != "singularity" ] && \
+[ ${container_software} != "podman" ]
+then
+    echo "Invalid container software ${container_software} given!"
+    printf "This script supports docker, singularity or podman only at this time.\n\n"
+    usage
+fi
+
+if ! command -v ${container_software} &> /dev/null
+then
+    echo "${container_software} executable could not be found. Please make sure ${container_software} is installed and available"
+    printf "in the PATH environment variable. Exiting.\n\n"
+    usage
+fi
+
+
 # infile
 function check_infile() {
     if [[ ! -f ${infile} ]]
@@ -119,9 +143,22 @@ function check_workdir() {
 # Run the docker image on $the prepared {workdir}
 function run_docker_image() {
     cp ${infile} ${workdir}/proteins.fasta
-    docker run \
-        --mount type=bind,source=${workdir},target=/volume \
-        dockerhub.ebi.ac.uk/uniprot-public/unifire:${docker_version}
+    if [ ${container_software} == "docker" ]
+    then
+      docker run \
+          --mount type=bind,source=${workdir},target=/volume \
+          dockerhub.ebi.ac.uk/uniprot-public/unifire:${docker_version}
+    elif [ ${container_software} == "singularity" ]
+    then
+      singularity run \
+          --bind ${workdir}:/volume \
+          docker://dockerhub.ebi.ac.uk/uniprot-public/unifire:${docker_version}
+    elif [ ${container_software} == "podman" ]
+    then
+      podman run \
+          --mount type=bind,source=${workdir},target=/volume \
+          docker://dockerhub.ebi.ac.uk/uniprot-public/unifire:${docker_version}
+    fi
 }
 
 # Move output files from ${workdir} to ${outdir}

docker/scripts/unifire-workflow.sh

@@ -18,12 +18,13 @@
 
 UNIFIRE_REPO="/opt/git/unifire"
 INTERPROSCAN_REPO="/opt/interproscan-5.45-80.0"
+ETE3FOLDER="/opt/ete3"
 
 VOLUME=/volume
 infilename=infile.fasta
 
-cd ${UNIFIRE_REPO}
-./misc/taxonomy/fetchLineageLocal.py ${VOLUME}/proteins.fasta ${VOLUME}/proteins_lineage.fasta
+${UNIFIRE_REPO}/misc/taxonomy/fetchTaxonomicLineage.py -i ${VOLUME}/proteins.fasta -o ${VOLUME}/proteins_lineage.fasta \
+  -t ${ETE3FOLDER}/taxa.sqlite
 
 ${INTERPROSCAN_REPO}/interproscan.sh -f xml -dp -i ${VOLUME}/proteins_lineage.fasta \
     --appl "Hamap,ProSiteProfiles,ProSitePatterns,Pfam,TIGRFAM,SMART,PRINTS,SFLD,CDD,Gene3D,ProDom,PIRSF,PANTHER,SUPERFAMILY" \

docker/scripts/update-taxonomy-cache.sh

+#!/usr/bin/env bash
+
+############################################################################
+#    Copyright (c) 2018 European Molecular Biology Laboratory
+#
+#    Licensed under the Apache License, Version 2.0 (the "License");
+#    you may not use this file except in compliance with the License.
+#    You may obtain a copy of the License at
+#
+#       http://www.apache.org/licenses/LICENSE-2.0
+#
+#    Unless required by applicable law or agreed to in writing, software
+#    distributed under the License is distributed on an "AS IS" BASIS,
+#    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#    See the License for the specific language governing permissions and
+#    limitations under the License.
+############################################################################
+
+set -e
+set -u
+
+SCRIPT_PATH=`dirname $0`
+ETE3FOLDER="/opt/ete3"
+
+${SCRIPT_PATH}/update-taxonomy-cache.py
+
+mkdir -p ${ETE3FOLDER}
+mv ~/.etetoolkit/taxa.sqlite ${ETE3FOLDER}/
+chmod 644 ${ETE3FOLDER}/taxa.sqlite

misc/taxonomy/fetchTaxonomicLineage.py

+#!/usr/bin/python
+# -*- coding: utf-8 -*-
+
+"""
+The fetchTaxonomicLineage.py script reads a MultiFasta file and will replace any occurence of "OX={taxId}" in the header
+by the full lineage corresponding to this taxId. The output is the resolved multifasta file written on the given
+output path.
+
+This script should be used when processing large amount of sequences from different species.
+
+All the taxonomy data from NCBI are stored locally via Ete (by default in ~/.etetoolkit/taxa.sqlite)
+This local storage can be updated using the following Python lines:
+
+    from ete3 import NCBITaxa
+    NCBITaxa().update_taxonomy_database()
+
+Library dependencies (via pip / conda / ...):
+    * ete3  (pip install ete3  /  conda install -c etetoolkit ete3)
+    * biopython
+"""
+import argparse
+
+from ete3 import NCBITaxa
+from Bio import SeqIO
+import sys, re
+
+__copyright__ = "Copyright 2018, European Molecular Biology Laboratory"
+__license__ = "Apache 2.0"
+__maintainer__ = "EMBL-EBI - Protein Function Development Team"
+__status__ = "Prototype"
+__author__ = "Alexandre Renaux"
+
+
+
+header_DE_remove_pattern = re.compile("([a-zA-Z0-9]+\|[a-zA-Z0-9]+\|[a-zA-Z0-9_]+)\s(.+?)(\s[A-Z]{2}=.+)")
+header_OX_pattern = re.compile('(OX=)(\d+)')
+
+taxId_to_lineage = {}
+
+
+def get_taxonomy_full_lineage(tax_id, ncbi):
+    if tax_id in taxId_to_lineage:
+        return taxId_to_lineage[tax_id]
+    else:
+        lineage = ncbi.get_lineage(tax_id)
+        taxId_to_lineage[tax_id] = lineage
+        return lineage
+
+
+def resolve_header(header, ncbi):
+    tax_id_match = header_OX_pattern.search(header)
+    if tax_id_match:
+        tax_id = int(tax_id_match.group(2))
+        lineage = get_taxonomy_full_lineage(tax_id, ncbi)
+        if lineage:
+            replacement = "\g<1>" + ",".join(str(i) for i in lineage)
+            return re.sub(header_OX_pattern, replacement, header)
+    return header
+
+
+def remove_long_protein_name(description):
+    match = header_DE_remove_pattern.search(description)
+    if match:
+        groups = list(header_DE_remove_pattern.search(description).groups())
+        if len(groups[1]) > 127:
+            del groups[1]
+        return " ".join(groups)
+    else:
+        return description
+
+
+def main(arguments):
+    file_in = arguments.infile
+    file_out = arguments.outfile
+    if arguments.taxadb is None:
+        ncbi = NCBITaxa()
+    else:
+        ncbi = NCBITaxa(dbfile=arguments.taxadb)
+    with open(file_out, 'w') as f_out:
+        for seq_record in SeqIO.parse(open(file_in, mode='r'), "fasta"):
+            seq_record.description = remove_long_protein_name(resolve_header(seq_record.description, ncbi))
+            seq_record.id = ""
+            r = SeqIO.write(seq_record, f_out, "fasta")
+            if r != 1: print("Error while writing sequence:  " + seq_record.id)
+
+def parse_args():
+    parser = argparse.ArgumentParser(description="""
+    The script fetchTaxonomicLineage.py reads an input file in multifasta format and will replace any occurence of 
+    OX={taxId}" in the header by the full lineage corresponding to this taxId.
+    """)
+    parser.add_argument('--infile', '-i', dest="infile", required=True, help="""
+    Path to the input file in multifasta format with one tax-id in each fasta header in the format OX={taxId}
+    """)
+    parser.add_argument('--outfile', '-o', dest="outfile", required=True, help="""
+    Path to the output file in multifasta format with the full taxonomic lineage in each fasta header in the format 
+    OX={taxId1,taxId2,...}
+    """)
+    parser.add_argument('--taxa-sqlite', '-t', dest="taxadb", required=False, help="""
+    Path to the sqlite DB file for taxonomy database. Default location is ~/.etetoolkit/taxa.sqlite
+    """)
+
+    return parser.parse_args()
+
+
+if __name__ == "__main__":
+    arguments = parse_args()
+    main(arguments)