Merge the dev branch

.github/workflows/ci.yml

@@ -8,19 +8,18 @@ jobs:
     strategy:
       fail-fast: true
       matrix:
-        python-version: ["3.8", "3.13"]
-        poetry-version: ["1.5.1"]
+        python-version: ["3.9", "3.13"]
         os: [ubuntu-22.04] #, macos-latest, windows-latest]
     runs-on: ${{ matrix.os }}
     steps:
-      - uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python-version }}
-      - name: Run image
-        uses: abatilo/actions-poetry@v2
-        with:
-          poetry-version: ${{ matrix.poetry-version }}
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install --upgrade poetry
       - name: Test build
         run: |
           poetry config virtualenvs.in-project true
@@ -32,22 +31,23 @@ jobs:
       fail-fast: true
     runs-on: ubuntu-22.04
     steps:
-      - uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v5
         with:
-          python-version: "3.11"
-      - name: Run image
-        uses: abatilo/actions-poetry@v2
-        with:
-          poetry-version: "1.5.1"
+          python-version: "3.13"
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install --upgrade poetry
+          poetry install --with dev
       - name: Install the project dependencies
         run: poetry install
       - name: Run black formatter
         run: poetry run black --check --exclude '__pycache__|scripts|sentieon_cli/archive.py' --line-length 79 sentieon_cli
       - name: Run mypy
-        run: poetry run mypy sentieon_cli
+        run: poetry run mypy sentieon_cli --exclude scripts
       - name: Run flake8
-        run: poetry run flake8 . --extend-ignore E231,E221, --exclude .github/scripts/license_message.py,sentieon_cli/scripts/gvcf_combine.py,sentieon_cli/scripts/vcf_mod.py,sentieon_cli/scripts/hybrid_anno.py,sentieon_cli/scripts/hybrid_select.py,sentieon_cli/archive.py,tests # false+ from python 3.12
+        run: poetry run flake8 . --extend-ignore E231,E221, --exclude .github/scripts/license_message.py,sentieon_cli/scripts/gvcf_combine.py,sentieon_cli/scripts/vcf_mod.py,sentieon_cli/scripts/hybrid_anno.py,sentieon_cli/scripts/hybrid_select.py,sentieon_cli/archive.py,sentieon_cli/scripts/trimalt.py,playground,tests # false+ from python 3.12
       - name: Run the automated tests
         run: poetry run pytest -v
       - name: Run doct tests

.github/workflows/main.yml

@@ -14,20 +14,19 @@ jobs:
       fail-fast: true
       max-parallel: 1
       matrix:
-        python-version: ["3.8", "3.13"]
-        poetry-version: ["1.5.1"]
-        sentieon-version: ["202503"]
+        python-version: ["3.9", "3.13"]
+        sentieon-version: ["202503.01"]
         os: [ubuntu-22.04] #, macos-latest, windows-latest]
     runs-on: ${{ matrix.os }}
     steps:
-      - uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python-version }}
-      - name: Run image
-        uses: abatilo/actions-poetry@v2
-        with:
-          poetry-version: ${{ matrix.poetry-version }}
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install --upgrade poetry
       - name: Install the project dependencies
         run: |
           poetry config virtualenvs.in-project true
@@ -42,9 +41,16 @@ jobs:
           pip install multiqc
       - name: Install sentieon
         run: |
-          curl -L https://s3.amazonaws.com/sentieon-release/software/sentieon-genomics-$SENTIEON_VERSION.tar.gz | tar -zxf -
+          url="https://s3.amazonaws.com/sentieon-release/software/sentieon-genomics-$SENTIEON_VERSION.tar.gz"
+          bin_dir=$(pwd)/sentieon-genomics-$SENTIEON_VERSION/bin
+          machine_arch=$(uname -m)
+          if [[ "$machine_arch" == "arm64" || "$machine_arch" == "aarch64" || "$machine_arch" == "armv7l" ]]; then
+            url="https://s3.amazonaws.com/sentieon-release/software/arm-sentieon-genomics-$SENTIEON_VERSION.tar.gz"
+            bin_dir=$(pwd)/arm-sentieon-genomics-$SENTIEON_VERSION/bin
+          fi
+          curl -L "$url" | tar -zxf -
           echo "SENTIEON_VERSION=$SENTIEON_VERSION" >> $GITHUB_ENV
-          echo "PATH=$(pwd)/sentieon-genomics-$SENTIEON_VERSION/bin:$PATH" >> $GITHUB_ENV
+          echo "PATH=$bin_dir:$PATH" >> $GITHUB_ENV
         env:
           SENTIEON_VERSION: ${{ matrix.sentieon-version }}
       - name: Download short-read model
@@ -91,28 +97,30 @@ jobs:
       fail-fast: true
       max-parallel: 1
       matrix:
-        python-version: ["3.8", "3.13"]
-        poetry-version: ["1.5.1"]
-        sentieon-version: ["202503"]
+        python-version: ["3.9", "3.13"]
+        sentieon-version: ["202503.01"]
         os: [ubuntu-22.04] #, macos-latest, windows-latest]
     runs-on: ${{ matrix.os }}
     steps:
-      - uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python-version }}
-      - name: Run image
-        uses: abatilo/actions-poetry@v2
-        with:
-          poetry-version: ${{ matrix.poetry-version }}
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install --upgrade poetry
       - name: Install the project dependencies
         run: |
           poetry config virtualenvs.in-project true
           poetry install
       - name: Install bcftools
         run: |
           sudo apt-get update
-          sudo apt-get install -y bcftools
+          sudo apt-get install libcurl4-gnutls-dev
+          curl -L https://github.com/samtools/bcftools/releases/download/1.22/bcftools-1.22.tar.bz2 | tar -jxf -
+          cd bcftools-1.22
+          ./configure && sudo make install
       - name: Install samtools
         run: |
           curl -L https://github.com/samtools/samtools/releases/download/1.19.2/samtools-1.19.2.tar.bz2 | tar -jxf -
@@ -128,9 +136,16 @@ jobs:
           sudo chmod ugo+x /usr/local/bin/mosdepth
       - name: Install sentieon
         run: |
-          curl -L https://s3.amazonaws.com/sentieon-release/software/sentieon-genomics-$SENTIEON_VERSION.tar.gz | tar -zxf -
+          url="https://s3.amazonaws.com/sentieon-release/software/sentieon-genomics-$SENTIEON_VERSION.tar.gz"
+          bin_dir=$(pwd)/sentieon-genomics-$SENTIEON_VERSION/bin
+          machine_arch=$(uname -m)
+          if [[ "$machine_arch" == "arm64" || "$machine_arch" == "aarch64" || "$machine_arch" == "armv7l" ]]; then
+            url="https://s3.amazonaws.com/sentieon-release/software/arm-sentieon-genomics-$SENTIEON_VERSION.tar.gz"
+            bin_dir=$(pwd)/arm-sentieon-genomics-$SENTIEON_VERSION/bin
+          fi
+          curl -L "$url" | tar -zxf -
           echo "SENTIEON_VERSION=$SENTIEON_VERSION" >> $GITHUB_ENV
-          echo "PATH=$(pwd)/sentieon-genomics-$SENTIEON_VERSION/bin:$PATH" >> $GITHUB_ENV
+          echo "PATH=$bin_dir:$PATH" >> $GITHUB_ENV
         env:
           SENTIEON_VERSION: ${{ matrix.sentieon-version }}
       - name: Download long-read model
@@ -204,28 +219,30 @@ jobs:
       fail-fast: true
       max-parallel: 1
       matrix:
-        python-version: ["3.8", "3.13"]
-        poetry-version: ["1.5.1"]
-        sentieon-version: ["202503.01.rc1"]
+        python-version: ["3.9", "3.13"]
+        sentieon-version: ["202503.01"]
         os: [ubuntu-22.04] #, macos-latest, windows-latest]
     runs-on: ${{ matrix.os }}
     steps:
-      - uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@v5
+      - uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python-version }}
-      - name: Run image
-        uses: abatilo/actions-poetry@v2
-        with:
-          poetry-version: ${{ matrix.poetry-version }}
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install --upgrade poetry
       - name: Install the project dependencies
         run: |
           poetry config virtualenvs.in-project true
           poetry install
       - name: Install bcftools
         run: |
           sudo apt-get update
-          sudo apt-get install -y bcftools
+          sudo apt-get install libcurl4-gnutls-dev
+          curl -L https://github.com/samtools/bcftools/releases/download/1.22/bcftools-1.22.tar.bz2 | tar -jxf -
+          cd bcftools-1.22
+          ./configure && sudo make install
       - name: Install samtools
         run: |
           curl -L https://github.com/samtools/samtools/releases/download/1.19.2/samtools-1.19.2.tar.bz2 | tar -jxf -
@@ -244,9 +261,16 @@ jobs:
           pip install multiqc
       - name: Install sentieon
         run: |
-          curl -L https://s3.amazonaws.com/sentieon-release/software/sentieon-genomics-$SENTIEON_VERSION.tar.gz | tar -zxf -
+          url="https://s3.amazonaws.com/sentieon-release/software/sentieon-genomics-$SENTIEON_VERSION.tar.gz"
+          bin_dir=$(pwd)/sentieon-genomics-$SENTIEON_VERSION/bin
+          machine_arch=$(uname -m)
+          if [[ "$machine_arch" == "arm64" || "$machine_arch" == "aarch64" || "$machine_arch" == "armv7l" ]]; then
+            url="https://s3.amazonaws.com/sentieon-release/software/arm-sentieon-genomics-$SENTIEON_VERSION.tar.gz"
+            bin_dir=$(pwd)/arm-sentieon-genomics-$SENTIEON_VERSION/bin
+          fi
+          curl -L "$url" | tar -zxf -
           echo "SENTIEON_VERSION=$SENTIEON_VERSION" >> $GITHUB_ENV
-          echo "PATH=$(pwd)/sentieon-genomics-$SENTIEON_VERSION/bin:$PATH" >> $GITHUB_ENV
+          echo "PATH=$bin_dir:$PATH" >> $GITHUB_ENV
         env:
           SENTIEON_VERSION: ${{ matrix.sentieon-version }}
       - name: Download hybrid model
@@ -294,3 +318,21 @@ jobs:
           if [ ! -f "output_hybrid2.vcf.gz" ]; then
             exit 1
           fi
+      - name: Smoke test - readgroups
+        run: |
+          . .venv/bin/activate
+          samtools addreplacerg -m overwrite_all \
+            -o "tests/smoke/illumina_2.bam" \
+            -r '@RG\tID:HG002-2\tSM:HG003\tPL:ILLUMINA' -O BAM \
+            --reference "tests/smoke/r ef.fa" "tests/smoke/illumina.cram"
+          samtools index -b "tests/smoke/illumina_2.bam"
+          sentieon-cli -v dnascope-hybrid -t 1 -r "tests/smoke/r ef.fa" \
+            --rgsm "HG001" \
+            --sr_aln "tests/smoke/illumina.cram" "tests/smoke/illumina_2.bam" \
+            --lr_aln "tests/smoke/sam ple.cram" \
+            -m "DNAscope hybrid IlmnPB1.1.bundle" \
+            --sr_duplicate_marking none \
+            "output_hybrid3.vcf.gz"
+          if [ ! -f "output_hybrid3.vcf.gz" ]; then
+            exit 1
+          fi

.github/workflows/release.yml

@@ -13,11 +13,11 @@ jobs:
       - uses: actions/checkout@v3
       - uses: actions/setup-python@v4
         with:
-          python-version: "3.12"
-      - name: Run image
-        uses: abatilo/actions-poetry@v2
-        with:
-          poetry-version: "1.5.1"
+          python-version: "3.13"
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install --upgrade poetry
       - name: Test build
         run: |
           poetry config virtualenvs.in-project true

README.md

@@ -8,8 +8,8 @@ A command-line interface for the Sentieon software
 
 Download the latest tar.gz file from the GitHub release page, https://github.com/sentieon/sentieon-cli/releases/ and install the package with pip:
 ```sh
-curl -LO https://github.com/Sentieon/sentieon-cli/releases/download/v1.4.0/sentieon_cli-1.4.0.tar.gz
-pip install sentieon_cli-1.4.0.tar.gz
+curl -LO https://github.com/Sentieon/sentieon-cli/releases/download/v1.5.0/sentieon_cli-1.5.0.tar.gz
+pip install sentieon_cli-1.5.0.tar.gz
 ```
 
 ## Installation with Poetry

docs/dnascope-hybrid.md

@@ -5,7 +5,7 @@ Sentieon DNAscope Hybrid is a pipeline for germline variant calling using combin
 The pipeline supports input data in the following formats; both short-read and long-read input are required:
 * Unaligned short-read data in gzipped FASTQ format.
 * Aligned short-reads in BAM or CRAM format.
-* Unaliged long-read data in the uBAM or uCRAM format.
+* Unaligned long-read data in the uBAM or uCRAM format.
 * Aligned long-read data in BAM or CRAM format.
 
 By default, the pipeline will generate the following output files:

docs/dnascope.md

@@ -94,7 +94,7 @@ sentieon-cli dnascope [-h] \
 ```
 
 With uBAM or uCRAM input, the DNAscope pipeline requires the following new arguments:
-- `-i SAMPLE_INPUT`: the input input sample file in uBAM or uCRAM format. One or more files can be supplied by passing multiple files after the `-i` argument.
+- `-i SAMPLE_INPUT`: the input sample file in uBAM or uCRAM format. One or more files can be supplied by passing multiple files after the `-i` argument.
 - `--align`: directs the pipeline to align the input reads.
 
 The DNAscope pipeline accepts the following new optional arguments: