Updates to the DNAscope LongRead pipeline #9
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
| name: CI-main | |
| on: | |
| push: | |
| branches: | |
| - main | |
| pull_request: | |
| jobs: | |
| smoke: | |
| strategy: | |
| fail-fast: true | |
| max-parallel: 1 | |
| matrix: | |
| python-version: ["3.8", "3.12"] | |
| poetry-version: ["1.5.1"] | |
| sentieon-version: ["202308.01", "202308.02"] | |
| os: [ubuntu-22.04] #, macos-latest, windows-latest] | |
| runs-on: ${{ matrix.os }} | |
| steps: | |
| - uses: actions/checkout@v3 | |
| - uses: actions/setup-python@v4 | |
| with: | |
| python-version: ${{ matrix.python-version }} | |
| - name: Run image | |
| uses: abatilo/actions-poetry@v2 | |
| with: | |
| poetry-version: ${{ matrix.poetry-version }} | |
| - name: Install the project dependencies | |
| run: | | |
| poetry config virtualenvs.in-project true | |
| poetry install | |
| - name: Install bcftools | |
| run: | | |
| sudo apt-get update | |
| sudo apt-get install -y bcftools | |
| - name: Install samtools | |
| run: | | |
| curl -L https://github.com/samtools/samtools/releases/download/1.19.2/samtools-1.19.2.tar.bz2 | tar -jxf - | |
| cd samtools-1.19.2 | |
| ./configure && sudo make install | |
| - name: Install bedtools | |
| run: | | |
| sudo curl -L -o /usr/local/bin/bedtools "https://github.com/arq5x/bedtools2/releases/download/v2.31.0/bedtools.static" | |
| sudo chmod ugo+x /usr/local/bin/bedtools | |
| - name: Install sentieon | |
| run: | | |
| curl -L https://s3.amazonaws.com/sentieon-release/software/sentieon-genomics-$SENTIEON_VERSION.tar.gz | tar -zxf - | |
| env: | |
| SENTIEON_VERSION: ${{ matrix.sentieon-version }} | |
| - name: Download model | |
| run: | | |
| curl -L "https://s3.amazonaws.com/sentieon-release/other/DNAscopePacBio2.1.bundle" \ | |
| > "DNAscope PacBio2.1.bundle" | |
| - name: Smoke test | |
| run: | | |
| SENTIEON_AUTH_DATA=$(python3 .github/scripts/license_message.py encrypt --key "$ENCRYPTION_KEY" --message "$LICENSE_MESSAGE") | |
| export SENTIEON_AUTH_DATA | |
| . .venv/bin/activate | |
| export PATH=$(pwd)/sentieon-genomics-$SENTIEON_VERSION/bin:$PATH | |
| gzip -dc tests/smoke/ref.fa.bgz > "tests/smoke/r ef.fa" | |
| mv tests/smoke/ref.fa.fai "tests/smoke/r ef.fa.fai" | |
| mv tests/smoke/sample.cram "tests/smoke/sam ple.cram" | |
| mv tests/smoke/sample.cram.crai "tests/smoke/sam ple.cram.crai" | |
| sentieon-cli dnascope-longread -t 1 -r "tests/smoke/r ef.fa" \ | |
| -i "tests/smoke/sam ple.cram" -m "DNAscope PacBio2.1.bundle" \ | |
| --repeat-model tests/smoke/sample_repeat.model -g \ | |
| -b "tests/smoke/diploid.bed" \ | |
| --haploid-bed "tests/smoke/haploid.bed" \ | |
| "output hifi.vcf.gz" | |
| sentieon driver -r "tests/smoke/r ef.fa" -t 1 --algo GVCFtyper \ | |
| -v "output hifi.g.vcf.gz" output_hifi_gvcftyper.vcf.gz | |
| if [ ! -f "output hifi.sv.vcf.gz" -o ! -f "output hifi.haploid.vcf.gz" ]; then | |
| exit 1 | |
| fi | |
| sentieon-cli dnascope-longread --tech ONT -t 1 -r "tests/smoke/r ef.fa" \ | |
| -i "tests/smoke/sam ple.cram" -m "DNAscope PacBio2.1.bundle" \ | |
| --repeat-model tests/smoke/sample_repeat.model -g "output ont.vcf.gz" | |
| sentieon driver -r "tests/smoke/r ef.fa" -t 1 --algo GVCFtyper \ | |
| -v "output ont.g.vcf.gz" output_hifi_gvcftyper.vcf.gz | |
| if [ ! -f "output ont.sv.vcf.gz" ]; then | |
| exit 1 | |
| fi | |
| sentieon-cli dnascope-longread -t 1 -r "tests/smoke/r ef.fa" \ | |
| -i "tests/smoke/sam ple.cram" -m "DNAscope PacBio2.1.bundle" \ | |
| --repeat-model tests/smoke/sample_repeat.model --align \ | |
| --input_ref "tests/smoke/r ef.fa" "output realigned.vcf.gz" | |
| if [ ! -f "output realigned.vcf.gz" -o ! -f "output realigned_mm2_sorted_0.cram" ]; then | |
| exit 1 | |
| fi | |
| samtools fastq --reference "tests/smoke/r ef.fa" \ | |
| "tests/smoke/sam ple.cram" | \ | |
| gzip -c > sample.fq.gz | |
| sentieon-cli dnascope-longread -t 1 -r "tests/smoke/r ef.fa" \ | |
| --fastq sample.fq.gz --readgroups '@RG\tID:sample-1\tSM:sample' \ | |
| -m "DNAscope PacBio2.1.bundle" \ | |
| --repeat-model tests/smoke/sample_repeat.model "output fq.vcf.gz" | |
| if [ ! -f "output fq.vcf.gz" -o ! -f "output fq_mm2_sorted_fq_0.cram" -o ! -f "output fq.sv.vcf.gz" ]; then | |
| exit 1 | |
| fi | |
| env: | |
| SENTIEON_LICENSE: ${{ secrets.SENTIEON_LICENSE }} | |
| SENTIEON_AUTH_MECH: "GitHub Actions - token" | |
| ENCRYPTION_KEY: ${{ secrets.ENCRYPTION_KEY }} | |
| LICENSE_MESSAGE: ${{ secrets.LICENSE_MESSAGE }} | |
| SENTIEON_VERSION: ${{ matrix.sentieon-version }} | |