This repo includes a two-stage Dockerfile that builds the spades conda
environment, installs the vendored GOTTCHA2 package, and copies the runtime
files into a slim Debian image.
From the repo root:
docker build -t spades-g2:local .Optional multi-platform publish examples:
docker buildx build --platform linux/arm64 -t spades-g2:[DATE]-arm64 .
docker buildx build --platform linux/amd64 -t spades-g2:[DATE]-amd64 .
docker buildx build --platform linux/amd64,linux/arm64 -t spades-g2:[DATE] -t spades-g2:latest .The image contains data/, scripts/, run_SPADES.sh, and post_install.sh.
Run the lightweight install checker before using a new image:
docker run --rm -it spades-g2:local ./scripts/check_install.shThe checker does not run the sequencing workflow. It verifies required command line tools, Python imports, SPADES data files, and Krona taxonomy wiring.
Mount your reads and database into the container so the pipeline can read inputs and write outputs back to the host filesystem.
docker run --rm -it \
-v "$PWD":/work \
-v /path/to/gottcha_db:/db:ro \
-w /work \
spades-g2:local \
/app/run_SPADES.sh \
-i /work/reads.fastq.gz \
-o /work/outdir \
-p sample \
-d /db/gottcha_db.species.fna \
-t 8 \
--spades-data /app/data \
--js-externalNotes:
-dexpects the GOTTCHA2 fast-profile database base path without extensions.- The database directory must contain:
<db>.syldb<db>.zip<db>.stats<db>.tax.tsv
- The image already contains
data/taxonomy_dbanddata/pathogen.tsv. - Add
--onttorun_SPADES.shfor long-read single-end inputs. - Use
--js-externalunless the HTML reports will be served from an environment that provides the local/publicdataJavaScript/CSS assets.
The runtime image does not bake in the full test/ directory. Mount the repo if
you want to run the bundled test script:
docker run --rm -it \
-v "$PWD":/work \
-w /work/test \
spades-g2:local \
./test_run.sh