Merge pull request #31 from samhorsfield96/docker

Add docker to CI

.github/workflows/docker_push.yml

@@ -0,0 +1,47 @@
+name: Build and push Docker image
+
+on:
+  push:
+    branches:
+      - 'master'
+      - 'docker'
+    tags:
+      - 'v*'
+  pull_request:
+    branches:
+      - 'master'
+  create:
+    tags:
+      - v*
+
+jobs:
+  docker-upload:
+    runs-on: ubuntu-latest
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v2
+    - name: Docker meta
+      id: meta
+      uses: docker/metadata-action@v4
+      with:
+        images: samhorsfield96/ggcaller
+    - name: Set up QEMU
+      uses: docker/setup-qemu-action@v1
+    - name: Set up Docker Buildx
+      uses: docker/setup-buildx-action@v1
+    - name: Login to DockerHub
+      uses: docker/login-action@v3
+      with:
+        username: ${{ secrets.DOCKER_REGISTRY_USERNAME }}
+        password: ${{ secrets.DOCKER_REGISTRY_PASSWORD }}
+    - name: Build and push
+      id: docker_build
+      uses: docker/build-push-action@v3
+      with:
+        push: ${{ github.event_name != 'pull_request' }}
+        tags: ${{ steps.meta.outputs.tags }}
+        labels: ${{ steps.meta.outputs.labels }}
+        file: docker/Dockerfile
+        provenance: false
+    - name: Image digest
+      run: echo ${{ steps.docker_build.outputs.digest }}

docker/Dockerfile

@@ -4,11 +4,27 @@ USER root
 
 # create a project directory inside user home
 ARG MAMBA_DOCKERFILE_ACTIVATE=1
-COPY . /src
-WORKDIR /src
 
+# create a project directory inside user home
+# (this isn't used with a clone running snakemake)
+ENV PROJECT_DIR $HOME/app
+RUN mkdir $PROJECT_DIR
+# copy the code in
+COPY . $PROJECT_DIR
+WORKDIR $PROJECT_DIR
+
+# build conda env
+ENV ENV_PREFIX $PROJECT_DIR/env
 COPY --chown=$user:$user docker/environment_docker.yml /tmp/environment_docker.yml
+
+COPY --chown=$user:$user docker/entrypoint.sh /usr/local/bin/
+RUN chmod u+x /usr/local/bin/entrypoint.sh
+
 RUN micromamba install -y -n base -f /tmp/environment_docker.yml && \
-    micromamba clean --all --yes && python -m pip install --no-deps --ignore-installed . \
-    && PATH=$PATH:/opt/conda/bin
-WORKDIR /workdir
+    micromamba clean --all --yes && \
+    python -m pip install --no-deps --ignore-installed . && \
+    mkdir ggc_db && \
+    ggcaller --balrog-db ggc_db && \
+    PATH=$PATH:/opt/conda/bin
+
+ENTRYPOINT [ "/usr/local/bin/entrypoint.sh" ]

docker/entrypoint.sh

@@ -0,0 +1,4 @@
+#!/bin/bash --login
+set -e
+
+exec "$@"

docs/installation.rst

@@ -17,11 +17,11 @@ First, install `Docker <https://docs.docker.com/get-docker/>`_ for your OS. If r
 
 To use the latest image, run::
 
-    docker pull samhorsfield96/ggcaller:latest
+    docker pull samhorsfield96/ggcaller:master
 
 To run ggCaller from the Docker Hub image, run::
 
-	cd test && docker run --rm -it -v $(pwd):/workdir samhorsfield96/ggcaller:latest ggcaller --refs pneumo_CL_group2.txt
+	cd test && docker run --rm -it -v $(pwd):/workdir -v $(pwd):/data samhorsfield96/ggcaller:master ggcaller --balrog-db /app/ggc_db --refs /workdir/pneumo_CL_group2_docker.txt --out /workdir/ggc_out
 
 You can also build the image yourself. First download and switch to the ggCaller repository::
 
@@ -33,7 +33,9 @@ Finally, build with Docker. This should take between 5-10 minutes to fully insta
 
 To run ggCaller from a local Docker build, run::
 
-	cd test && docker run --rm -it -v $(pwd):/workdir ggc_env:latest ggcaller --refs pneumo_CL_group2.txt
+	cd test && docker run --rm -it -v $(pwd):/workdir -v $(pwd):/data ggc_env:latest ggcaller --balrog-db /app/ggc_db --refs /workdir/pneumo_CL_group2_docker.txt --out /workdir/ggc_out
+
+Please ensure you keep ``--balrog-db /app/ggc_db`` and ``/workdir`` paths as specified above.
 
 Installing with singularity
 -----------------------------------

docs/quickstart.rst

@@ -15,13 +15,17 @@ The easiest way to get up and running is using Docker. To get up and running, pu
 Preparing the data
 ------------------
 
-Place all of you samples to analyse in the same directory. Then navigate inside and run::
+Place all of your samples to be analysed in the same directory. Then navigate inside and run::
 
     ls -d -1 $PWD/*.fasta > input.txt
 
 If using Docker, instead navigate to the directory containing the fasta files and run the below command, to ensure file paths are relative (the docker version will not work with absolute paths)::
 
-    ls -d -1 *.fasta > input.txt
+    ls -d -1 *.fasta > input_docker.txt
+
+Then, append the prefix ``/data/`` to each line to enable ggCaller to find the files::
+
+    sed -i -e 's|^|/data/|' input_docker.txt
 
 Running ggCaller
 ------------------
@@ -39,9 +43,9 @@ To run ggCaller with just reads::
 
     ggcaller --reads input.txt --out output_path
 
-If using Docker, run with the below command. You must ensure all paths are relative, including in ``input.txt``::
+If using Docker, run with the below command, ensuring you keep ``--balrog-db /app/ggc_db`` and ``/workdir`` paths as specified below. Replace ``path to files`` with the absolute path to the directory of files in ``input_docker.txt``::
 
-	docker run --rm -it -v $(pwd):/workdir samhorsfield96/ggcaller:latest ggcaller --refs input.txt --out output_path
+	docker run --rm -it -v $(pwd):/workdir -v <path to files>:/data samhorsfield96/ggcaller:master ggcaller --balrog-db /app/ggc_db --refs /workdir/input_docker.txt --out /workdir/output_path 
 
 .. important::
     We haven't extensively tested calling genes within

ggCaller/__init__.py

@@ -2,4 +2,4 @@
 
 '''ggCaller: a gene caller for Bifrost graphs'''
 
-__version__ = '1.3.4'
+__version__ = '1.3.5'

ggCaller/__main__.py

@@ -326,6 +326,12 @@ def get_options():
 
     # Other options
     Misc = parser.add_argument_group('Misc. options')
+    Misc.add_argument("--balrog-db",
+                      default=None,
+                      dest="balrog_db",
+                      help="Path to save BALROG and default annotation databases. If not specified will download"
+                            "automatically on first run."
+                            "[Default = None]")
     Misc.add_argument("--quiet",
                       dest="verbose",
                       help="suppress additional output"
@@ -420,20 +426,24 @@ def main():
             options.reads is not None) and (options.query is None):
         graph_tuple = graph.build(options.refs, options.kmer, stop_codons_for, stop_codons_rev, start_codons_for,
                                   start_codons_rev, options.threads, False, options.no_write_graph, options.reads, ref_set)
+    elif options.balrog_db is not None:
+        db_dir = download_db(options.balrog_db)
+        sys.exit(0)
     else:
         print("Error: incorrect number of input files specified. Please only specify the below combinations:\n"
               "- Bifrost GFA and Bifrost colours file (with/without list of reference files)\n"
               "- Bifrost GFA, Bifrost colours file and list of query sequences\n"
               "- List of reference files\n"
               "- List of read files\n"
-              "- A list of reference files and a list of read files.")
+              "- A list of reference files and a list of read files.\n"
+              "- A path to download the balrog gene model files.")
         sys.exit(1)
 
     # unpack ORF pair into overlap dictionary and list for gene scoring
     input_colours, nb_colours, overlap, ref_list = graph_tuple
 
     # download balrog and annotation files
-    db_dir = download_db()
+    db_dir = download_db(options.balrog_db)
 
     # set rest of panaroo arguments
     options = set_default_args(options, nb_colours)
@@ -486,7 +496,7 @@ def main():
     # load models models if required
     if not options.no_filter:
         print("Loading gene models...")
-        ORF_model_file, TIS_model_file = load_balrog_models()
+        ORF_model_file, TIS_model_file = load_balrog_models(db_dir)
 
     else:
         ORF_model_file, TIS_model_file = "NA", "NA"

models/__main__.py

@@ -4,28 +4,39 @@
 
 """ Get directories for model and seengenes """
 module_dir = os.path.dirname(os.path.realpath(__file__))
-zipped_db_dir = db_dir = os.path.join(module_dir, "ggCallerdb.tar.bz2")
-db_dir = os.path.join(module_dir, "ggCallerdb")
-balrog_model_dir = os.path.join(db_dir, "balrog_models")
+module_zipped_db_dir = os.path.join(module_dir, "ggCallerdb.tar.bz2")
+module_db_dir = os.path.join(module_dir, "ggCallerdb")
+module_balrog_model_dir = os.path.join(module_db_dir, "balrog_models")
 
-def download_db():
-    if not os.path.exists(zipped_db_dir):
+def download_db(download_db=None):
+    if download_db is None:
+        zipped_db_path = module_zipped_db_dir
+        db_path = module_db_dir
+        output_dir = module_dir
+    else:
+        zipped_db_path = os.path.join(download_db, "ggCallerdb.tar.bz2")
+        db_path = os.path.join(download_db, "ggCallerdb")
+        output_dir = download_db
+
+    if not os.path.exists(zipped_db_path):
         print("Downloading databases...")
         url = "https://ftp.ebi.ac.uk/pub/databases/pp_dbs/ggCallerdb.tar.bz2"
-        filename = wget.download(url, out=module_dir)
+        filename = wget.download(url, out=output_dir)
         print("")
-    if not os.path.exists(db_dir):
-        tar = tarfile.open(db_dir + ".tar.bz2", mode="r:bz2")
-        tar.extractall(module_dir)
+    if not os.path.exists(db_path):
+        tar = tarfile.open(zipped_db_path, mode="r:bz2")
+        tar.extractall(output_dir)
         tar.close()
 
-    return db_dir
+    return db_path
 
-def load_balrog_models():
+def load_balrog_models(db_path):
+    balrog_model_dir = os.path.join(db_path, "balrog_models")
+
     # check if directory exists. If not, unzip file
     if not os.path.exists(balrog_model_dir):
         tar = tarfile.open(balrog_model_dir + ".tar.gz", mode="r:gz")
-        tar.extractall(db_dir)
+        tar.extractall(db_path)
         tar.close()
 
     geneTCN = os.path.join(balrog_model_dir, "geneTCN_jit.pt")

test/pneumo_CL_group2_docker.txt

@@ -0,0 +1,5 @@
+/data/CR931658_Streptococcus_pneumoniae_strain_559_66_serotype_12a.fa
+/data/CR931659_Streptococcus_pneumoniae_strain_Gambia_1_81_serotype_12b.fa
+/data/CR931660_Streptococcus_pneumoniae_strain_6312_serotype_12f.fa
+/data/CR931717_Streptococcus_pneumoniae_strain_Hammer_serotype_44.fa
+/data/CR931719_Streptococcus_pneumoniae_strain_Eddy_nr._73_serotype_46.fa