-
Notifications
You must be signed in to change notification settings - Fork 125
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #1110 from StaPH-B/cjk-pdata-1.31
adds pangolin 4.3.1 and pangolin-data 1.31
- Loading branch information
Showing
3 changed files
with
185 additions
and
1 deletion.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,131 @@ | ||
| FROM mambaorg/micromamba:2.0.3-ubuntu22.04 AS app | ||
|
|
||
| # build and run as root users since micromamba image has 'mambauser' set as the $USER | ||
| USER root | ||
| # set workdir to default for building; set to /data at the end | ||
| WORKDIR / | ||
|
|
||
| # ARG variables only persist during build time | ||
| # had to include the v for some of these due to GitHub tags. | ||
| # using pangolin-data github tag, NOT what is in the GH release title "v1.2.133" | ||
| ARG PANGOLIN_VER="v4.3.1" | ||
| ARG PANGOLIN_DATA_VER="v1.31" | ||
| ARG SCORPIO_VER="v0.3.19" | ||
| ARG CONSTELLATIONS_VER="v0.1.12" | ||
| ARG USHER_VER="0.6.3" | ||
|
|
||
| # metadata labels | ||
| LABEL base.image="mambaorg/micromamba:2.0.3-ubuntu22.04" | ||
| LABEL dockerfile.version="1" | ||
| LABEL software="pangolin" | ||
| LABEL software.version=${PANGOLIN_VER} | ||
| LABEL description="Conda environment for Pangolin. Pangolin: Software package for assigning SARS-CoV-2 genome sequences to global lineages." | ||
| LABEL website="https://github.com/cov-lineages/pangolin" | ||
| LABEL license="GNU General Public License v3.0" | ||
| LABEL license.url="https://github.com/cov-lineages/pangolin/blob/master/LICENSE.txt" | ||
| LABEL maintainer="Curtis Kapsak" | ||
| LABEL maintainer.email="[email protected]" | ||
|
|
||
| # install dependencies; cleanup apt garbage | ||
| RUN apt-get update && apt-get install -y --no-install-recommends \ | ||
| wget \ | ||
| ca-certificates \ | ||
| git \ | ||
| procps \ | ||
| bsdmainutils && \ | ||
| apt-get autoclean && rm -rf /var/lib/apt/lists/* | ||
|
|
||
| # get the pangolin repo | ||
| RUN wget "https://github.com/cov-lineages/pangolin/archive/${PANGOLIN_VER}.tar.gz" && \ | ||
| tar -xf ${PANGOLIN_VER}.tar.gz && \ | ||
| rm -v ${PANGOLIN_VER}.tar.gz && \ | ||
| mv -v pangolin-* pangolin | ||
|
|
||
| # set the environment; PATH is unnecessary here, but leaving anyways. It's reset later in dockerfile | ||
| ENV PATH="$PATH" \ | ||
| LC_ALL=C.UTF-8 | ||
|
|
||
| # modify environment.yml to pin specific versions during install | ||
| # pin specific versions of usher, scorpio, pangolin-data, constellations, and pulp | ||
| # create the conda environment using modified environment.yml | ||
| # line to remove "defaults" channel to ensure that it isn't used due to Anaconda's recent ToS changes | ||
| RUN sed -i "s|usher.*|usher=${USHER_VER}|" /pangolin/environment.yml && \ | ||
| sed -i "s|scorpio.git|scorpio.git@${SCORPIO_VER}|" /pangolin/environment.yml && \ | ||
| sed -i "s|pangolin-data.git|pangolin-data.git@${PANGOLIN_DATA_VER}|" /pangolin/environment.yml && \ | ||
| sed -i "s|constellations.git|constellations.git@${CONSTELLATIONS_VER}|" /pangolin/environment.yml && \ | ||
| sed -i "12 a\ - pulp=2.7.0" /pangolin/environment.yml && \ | ||
| sed -i '/.*defaults/d' /pangolin/environment.yml && \ | ||
| micromamba create -n pangolin -y -f /pangolin/environment.yml && \ | ||
| micromamba clean -a -y -f | ||
|
|
||
| # so that mamba/conda env is active when running below commands | ||
| ENV ENV_NAME="pangolin" | ||
| ARG MAMBA_DOCKERFILE_ACTIVATE=1 | ||
|
|
||
| WORKDIR /pangolin | ||
|
|
||
| # run pip install step; download optional pre-computed assignment hashes for UShER (useful for running on large batches of samples) | ||
| # best to skip using the assigment-cache if running on one sample for speed | ||
| # print versions | ||
| RUN pip install . && \ | ||
| pangolin --add-assignment-cache && \ | ||
| mkdir /data && \ | ||
| pangolin --all-versions && \ | ||
| usher --version | ||
|
|
||
| # final working directory in "app" layer is /data for passing data in/out of container | ||
| WORKDIR /data | ||
|
|
||
| # hardcode pangolin executable into the PATH variable | ||
| ENV PATH="${PATH}:/opt/conda/envs/pangolin/bin/" XDG_CACHE_HOME=/tmp | ||
|
|
||
| # default command is to pull up help options for pangolin; can be overridden of course | ||
| CMD ["pangolin", "-h"] | ||
|
|
||
| # new base for testing | ||
| FROM app AS test | ||
|
|
||
| # so that mamba/conda env is active when running below commands | ||
| ENV ENV_NAME="pangolin" | ||
| ARG MAMBA_DOCKERFILE_ACTIVATE=1 | ||
|
|
||
| # test on test sequences supplied with Pangolin code | ||
| RUN pangolin /pangolin/pangolin/test/test_seqs.fasta -o /data/test_seqs-output-pusher && \ | ||
| column -t -s, /data/test_seqs-output-pusher/lineage_report.csv | ||
|
|
||
| # test functionality of assignment-cache option | ||
| RUN pangolin --use-assignment-cache /pangolin/pangolin/test/test_seqs.fasta | ||
|
|
||
| # download B.1.1.7 genome from Utah | ||
| ADD https://raw.githubusercontent.com/StaPH-B/docker-builds/master/tests/SARS-CoV-2/SRR13957123.consensus.fa /test-data/SRR13957123.consensus.fa | ||
|
|
||
| # test on a B.1.1.7 genome | ||
| RUN pangolin /test-data/SRR13957123.consensus.fa -o /test-data/SRR13957123-pusher && \ | ||
| column -t -s, /test-data/SRR13957123-pusher/lineage_report.csv | ||
|
|
||
| # install unzip for unzipping zip archive from NCBI | ||
| RUN apt-get update && apt-get install -y --no-install-recommends unzip | ||
|
|
||
| # install ncbi datasets tool (pre-compiled binary); place in $PATH | ||
| RUN wget https://ftp.ncbi.nlm.nih.gov/pub/datasets/command-line/LATEST/linux-amd64/datasets && \ | ||
| chmod +x datasets && \ | ||
| mv -v datasets /usr/local/bin | ||
|
|
||
| # testing the following lineages: | ||
| # BA.1 | ON924087.1 | from Florida https://www.ncbi.nlm.nih.gov/biosample?term=SAMN29506515 and https://www.ncbi.nlm.nih.gov/nuccore/ON924087 | ||
| # XBB.1.16 | OQ381818.1 | introduced in p-data 1.19, https://www.ncbi.nlm.nih.gov/nuccore/2440446687 and https://www.ncbi.nlm.nih.gov/biosample?term=SAMN33060589 and https://github.com/cov-lineages/pango-designation/issues/1723 | ||
| # another XBB.1.16 | OR177999.1 | https://www.ncbi.nlm.nih.gov/nuccore/OR177999.1 | ||
| # BA.2.86 | OR461132.1 | from Michigan https://www.ncbi.nlm.nih.gov/nuccore/OR461132.1 | ||
| # JN.2 (BA.2.86 sublineage) JN.2 is an alias of B.1.1.529.2.86.1.2 | OR598183.1 | NY CDC Quest sample https://www.ncbi.nlm.nih.gov/nuccore/OR598183 | ||
| # JQ.1 (BA.2.86.3 sublineage); JQ.1 is an alias of B.1.1.529.2.86.3.1 | OR716684.1 | THANK YOU ERIN AND UPHL!! https://www.ncbi.nlm.nih.gov/nuccore/OR716684 this test is important due to the fact that this lineage was included in the UShER tree, despite being designated after the pangolin-designation 1.23 release it previously caused and error/bug in pangolin, but now is fixed | ||
| # JN.1.22 (BA.2.86.x sublineage; full unaliased lineage is B.1.1.529.2.86.1.1.22) | PP189069.1 | https://github.com/cov-lineages/pango-designation/commit/a90c8e31c154621ed86c985debfea09e17541cda | ||
| # JN.1.48 (BA.2.86.x sublineage; full unaliased lineage is B.1.1.529.2.86.1.1.48) | PP218754.1 | https://github.com/cov-lineages/pango-designation/releases/tag/v1.27 and https://github.com/cov-lineages/pango-designation/commit/67f48bf24283999f1940f3aee8159f404124ff3f and https://www.ncbi.nlm.nih.gov/nuccore/PP218754 | ||
| # LK.1 | PP770375.1 | introduced in pango-designation 1.28 https://github.com/cov-lineages/pango-designation/commit/922795c90de355e67200cf4d379e8e5ff22472e4 and https://www.ncbi.nlm.nih.gov/nuccore/2728145425 thank you Luis, Lorraine, Marcos & team from PR Sci Trust for sharing your data! | ||
| # KP.3.3.2 | PQ073669.1 | introduced in pango-designation 1.29 https://github.com/cov-lineages/pango-designation/commit/7125e606818312b78f0756d7fcab6dba92dd0a9e and https://www.ncbi.nlm.nih.gov/nuccore/PQ073669 | ||
| # MC.2 | PQ034842.1 | introduced in pango-designation 1.30 https://github.com/cov-lineages/pango-designation/commit/c64dbc47fbfbfd7f4da011deeb1a88dd6baa45f1#diff-a121ea4b8cbeb4c0020511b5535bf24489f0223cc83511df7b8209953115d329R2564181 and https://www.ncbi.nlm.nih.gov/nuccore/PQ034842 | ||
| # XEC.3 | PQ277908.1 | introduced in pango-designation 1.31 https://github.com/cov-lineages/pango-designation/commit/ba3711a5615956ed97150288eb68356aa0fe7cdd#diff-a121ea4b8cbeb4c0020511b5535bf24489f0223cc83511df7b8209953115d329R2572545 and https://www.ncbi.nlm.nih.gov/nuccore/PQ277908.1 | ||
| RUN datasets download virus genome accession ON924087.1,OQ381818.1,OR177999.1,OR461132.1,OR598183.1,OR716684.1,PP189069.1,PP218754.1,PP770375.1,PQ073669.1,PQ034842.1,PQ277908.1 && \ | ||
| unzip -o ncbi_dataset.zip && \ | ||
| rm -v ncbi_dataset.zip && \ | ||
| pangolin ncbi_dataset/data/genomic.fna && \ | ||
| column -t -s, lineage_report.csv |
Oops, something went wrong.