Add workflow to detect structural variants

.github/workflows/build-hic-breakfinder-dockerfile.yml

@@ -0,0 +1,29 @@
+# Copyright (c) 2023 Roberto Rossini ([email protected])
+# SPDX-License-Identifier: MIT
+
+name: Build hic_breakfinder Dockerfile
+
+on:
+  push:
+    branches: [ main ]
+    paths:
+      - ".github/workflows/build-hic-breakfinder*dockerfile.yml"
+      - ".github/workflows/build-dockerfile.yml"
+      - "containers/hic-breakfinder*.Dockerfile"
+  pull_request:
+    branches: [ main ]
+    paths:
+      - ".github/workflows/build-hic-breakfinder*dockerfile.yml"
+      - ".github/workflows/build-dockerfile.yml"
+      - "containers/hic-breakfinder*.Dockerfile"
+
+jobs:
+  build-hic-breakfinder-dockerfile:
+    name: Build hic_breakfinder Dockerfile
+    uses: paulsengroup/2022-mcf10a-cancer-progression/.github/workflows/build-dockerfile.yml@main
+    with:
+      dockerfile-glob: "containers/hic-breakfinder*.Dockerfile"
+
+    permissions:
+      contents: read
+      packages: write

.github/workflows/build-hictrans-dockerfile.yml

@@ -0,0 +1,29 @@
+# Copyright (c) 2023 Roberto Rossini ([email protected])
+# SPDX-License-Identifier: MIT
+
+name: Build HiCTrans Dockerfile
+
+on:
+  push:
+    branches: [ main ]
+    paths:
+      - ".github/workflows/build-hictrans*dockerfile.yml"
+      - ".github/workflows/build-dockerfile.yml"
+      - "containers/hictrans*.Dockerfile"
+  pull_request:
+    branches: [ main ]
+    paths:
+      - ".github/workflows/build-hictrans*dockerfile.yml"
+      - ".github/workflows/build-dockerfile.yml"
+      - "containers/hictrans*.Dockerfile"
+
+jobs:
+  build-hictrans-dockerfile:
+    name: Build HiCTrans Dockerfile
+    uses: paulsengroup/2022-mcf10a-cancer-progression/.github/workflows/build-dockerfile.yml@main
+    with:
+      dockerfile-glob: "containers/hictrans*.Dockerfile"
+
+    permissions:
+      contents: read
+      packages: write

.github/workflows/build-hint-dockerfile.yml

@@ -0,0 +1,31 @@
+# Copyright (c) 2023 Roberto Rossini ([email protected])
+# SPDX-License-Identifier: MIT
+
+name: Build HiNT Dockerfile
+
+on:
+  push:
+    branches: [ main ]
+    paths:
+      - ".github/workflows/build-hint*dockerfile.yml"
+      - ".github/workflows/build-dockerfile.yml"
+      - "containers/hint*.Dockerfile"
+      - "containers/patches/hint.patch"
+  pull_request:
+    branches: [ main ]
+    paths:
+      - ".github/workflows/build-hint*dockerfile.yml"
+      - ".github/workflows/build-dockerfile.yml"
+      - "containers/hint*.Dockerfile"
+      - "containers/patches/hint.patch"
+
+jobs:
+  build-hint-dockerfile:
+    name: Build  HiNT Dockerfile
+    uses: paulsengroup/2022-mcf10a-cancer-progression/.github/workflows/build-dockerfile.yml@main
+    with:
+      dockerfile-glob: "containers/hint*.Dockerfile"
+
+    permissions:
+      contents: read
+      packages: write

bin/compute_restriction_sites_for_hint.py

@@ -0,0 +1,35 @@
+#!/usr/bin/env python3
+
+# Copyright (C) 2023 Roberto Rossini <[email protected]>
+#
+# SPDX-License-Identifier: MIT
+
+import argparse
+
+import bioframe as bf
+import pandas as pd
+
+
+def make_cli():
+    cli = argparse.ArgumentParser()
+
+    cli.add_argument("fna", type=str)
+    cli.add_argument("enzymes", nargs="+", type=str)
+
+    return cli
+
+
+if __name__ == "__main__":
+    args = vars(make_cli().parse_args())
+
+    dfs = []
+
+    fna = bf.load_fasta(args["fna"])
+    for enz in args["enzymes"]:
+        dfs.append(bf.digest(fna, enz))
+
+    df = pd.concat(dfs)[["chrom", "start"]].drop_duplicates(keep="first")
+
+    groups = df.groupby("chrom")["start"].aggregate(list)
+    for chrom, sites in groups.apply(lambda l: " ".join(str(x) for x in sorted(l))).items():
+        print(chrom, sites)

configs/detect_structural_variants.config

@@ -0,0 +1,57 @@
+// Copyright (C) 2022 Roberto Rossini <[email protected]>
+//
+// SPDX-License-Identifier: MIT
+
+params {
+    data_dir = 'data'
+    input_dir = "${data_dir}/input"
+    output_dir = "${data_dir}/output/structural_variants"
+
+    mcools = "${data_dir}/output/nfcore_hic/mcools/hg38*_merged.mcool"
+    nfcore_hic_samplesheet = "${data_dir}/input/nfcore_hic_samplesheet.csv"
+    alignment_dir = "${data_dir}/output/nfcore_hic/alignments"
+
+    ref_genome_name = 'hg38'
+    reference_genome = "${data_dir}/input/${ref_genome_name}/${ref_genome_name}.filtered.fa"
+    blacklist = "${data_dir}/input/${ref_genome_name}/${ref_genome_name}_blacklist.bed.gz"
+
+    hictrans_resolution = 10000
+
+    hint_resolution = 50  // kb
+    hint_refzip = "${data_dir}/input/hint/refData_${ref_genome_name}.zip"
+    hint_backdirzip = "${data_dir}/input/hint/backgroundMatrices_${ref_genome_name}.zip"
+    restriction_enzymes = 'DpnII HindIII'
+    restriction_enzymes_alias = 'arima'
+
+    hic_breakfinder_quality_score = 30
+    hic_breakfinder_expected_intra = "${data_dir}/input/hic_breakfinder/intra_expect_100kb.${ref_genome_name}.txt"
+    hic_breakfinder_expected_inter = "${data_dir}/input/hic_breakfinder/inter_expect_1Mb.${ref_genome_name}.txt"
+}
+
+process {
+    container = 'ghcr.io/paulsengroup/2022-mcf10a-cancer-progression/hictrans:e26ad6a'
+    withName:run_hictrans {
+        memory = 40.GB
+    }
+    withName:digest_genome_for_hint {
+        container = 'ghcr.io/paulsengroup/2022-mcf10a-cancer-progression/hint:2.2.8'
+    }
+    withName:run_hint_cnv {
+        memory = 6.GB
+        container = 'ghcr.io/paulsengroup/2022-mcf10a-cancer-progression/hint:2.2.8'
+    }
+    withName:run_hint_tl {
+        memory = 40.GB
+        container = 'ghcr.io/paulsengroup/2022-mcf10a-cancer-progression/hint:2.2.8'
+    }
+    withName:filter_mappings {
+        container = 'ghcr.io/paulsengroup/2022-mcf10a-cancer-progression/samtools:1.17'
+    }
+    withName:merge_bams {
+        container = 'ghcr.io/paulsengroup/2022-mcf10a-cancer-progression/samtools:1.17'
+    }
+    withName:run_hic_breakfinder {
+        memory = 165.GB
+        container = 'ghcr.io/paulsengroup/2022-mcf10a-cancer-progression/hic-breakfinder:30a0dcc'
+    }
+}

containers/hic-breakfinder__v30a0dcc.Dockerfile

@@ -0,0 +1,71 @@
+# Copyright (C) 2023 Roberto Rossini <[email protected]>
+#
+# SPDX-License-Identifier: MIT
+
+FROM mambaorg/micromamba:1.4.3 AS builder
+
+ARG CONTAINER_VERSION
+
+RUN if [ -z "$CONTAINER_VERSION" ]; then echo "Missing CONTAINER_VERSION --build-arg" && exit 1; fi
+
+ARG MAMBA_DOCKERFILE_ACTIVATE=1
+
+ARG HIC_BREAKFINDER_GIT='https://github.com/dixonlab/hic_breakfinder.git'
+ARG HIC_BREAKFINDER_TAG="${CONTAINER_VERSION}"
+
+RUN micromamba install -y \
+    -c conda-forge \
+    -c bioconda \
+    bamtools \
+    eigen \
+    git \
+    gxx \
+    make
+
+RUN cd /tmp \
+&& git clone "$HIC_BREAKFINDER_GIT" \
+&& cd hic_breakfinder \
+&& git checkout "$HIC_BREAKFINDER_TAG"
+
+RUN cd hic_breakfinder \
+&& CXXFLAGS='-isystem /opt/conda/include/bamtools -isystem /opt/conda/include/eigen3 -fpermissive' \
+  ./configure \
+&& make -j $(nproc)
+
+
+FROM mambaorg/micromamba:1.4.3 AS base
+
+ARG CONTAINER_VERSION
+
+RUN if [ -z "$CONTAINER_VERSION" ]; then echo "Missing CONTAINER_VERSION --build-arg" && exit 1; fi
+
+ARG MAMBA_DOCKERFILE_ACTIVATE=1
+ARG HINT_VERSION="${CONTAINER_VERSION}"
+
+
+RUN micromamba install -y \
+    -c conda-forge \
+    -c bioconda \
+    bamtools \
+    procps-ng \
+&& micromamba clean --all -y
+
+COPY --from=builder --chown=nobody:nogroup /tmp/hic_breakfinder/src/hic_breakfinder /usr/local/bin/
+COPY --from=builder --chown=nobody:nogroup /tmp/hic_breakfinder/LICENSE /usr/local/share/licenses/hic_breakfinder/
+
+WORKDIR /data
+
+ENV PATH="/opt/conda/bin:$PATH"
+ENTRYPOINT ["/usr/local/bin/_entrypoint.sh"]
+CMD ["/usr/local/bin/hic_breakfinder"]
+WORKDIR /data
+
+RUN whereis hic_breakfinder
+
+LABEL org.opencontainers.image.authors='Roberto Rossini <[email protected]>'
+LABEL org.opencontainers.image.url='https://github.com/paulsengroup/2022-mcf10a-cancer-progression'
+LABEL org.opencontainers.image.documentation='https://github.com/paulsengroup/2022-mcf10a-cancer-progression'
+LABEL org.opencontainers.image.source='https://github.com/paulsengroup/2022-mcf10a-cancer-progression'
+LABEL org.opencontainers.image.licenses='MIT'
+LABEL org.opencontainers.image.title="${CONTAINER_TITLE:-hic_breakfinder}"
+LABEL org.opencontainers.image.version="${CONTAINER_VERSION:-latest}"

containers/hictrans__ve26ad6a.Dockerfile

@@ -0,0 +1,83 @@
+# Copyright (C) 2023 Roberto Rossini <[email protected]>
+#
+# SPDX-License-Identifier: MIT
+
+FROM ubuntu:22.04 AS download
+
+ARG CONTAINER_VERSION
+ARG HICTRANS_VERSION="${CONTAINER_VERSION}"
+
+RUN apt-get update \
+&& apt-get install -y git \
+&& cd /tmp \
+&& git clone https://github.com/ay-lab/HiCtrans.git \
+&& cd HiCtrans \
+&& git checkout "$HICTRANS_VERSION"
+
+
+FROM mambaorg/micromamba:1.4.3 AS base
+
+ARG CONTAINER_VERSION
+
+RUN if [ -z "$CONTAINER_VERSION" ]; then echo "Missing CONTAINER_VERSION --build-arg" && exit 1; fi
+
+ARG MAMBA_DOCKERFILE_ACTIVATE=1
+ARG HICTRANS_VERSION="${CONTAINER_VERSION}"
+
+
+RUN micromamba install -y \
+    -c conda-forge \
+    -c bioconda \
+    cooler \
+    pigz \
+    procps-ng \
+    r-catools \
+    r-changepoint \
+    r-data.table \
+    r-depmixs4 \
+    r-deoptimr \
+    r-hashmap \
+    r-optparse \
+    r-r.utils \
+    r-rcpp \
+&& micromamba clean --all -y
+
+COPY --from=download --chown=nobody:nogroup /tmp/HiCtrans/hictrans.v3.R /opt/hictrans/
+
+USER root
+RUN mkdir /opt/hictrans/bin \
+&& printf '%s\n%s "$@"' \
+    '#!/usr/bin/env sh' \
+    'Rscript /opt/hictrans/hictrans.v3.R' > /opt/hictrans/bin/hictrans \
+&& chown -R nobody:nogroup /opt/hictrans \
+&& chmod 755 /opt/hictrans/bin/hictrans
+USER mambauser
+
+WORKDIR /data
+
+ENV PATH="/opt/conda/bin:/opt/hictrans/bin:$PATH"
+ENTRYPOINT ["/usr/local/bin/_entrypoint.sh"]
+CMD ["/opt/hictrans/bin/hictrans"]
+WORKDIR /data
+
+# We have to explicitly set these R_* env variables in order for the
+# container to work correctly when running using Apptainer
+ENV R_HOME=/opt/conda/lib/R
+ENV R_LIBS=/opt/conda/lib/R/lib
+ENV R_ENVIRON=/opt/conda/lib/R/etc/Renviron
+ENV R_HISTFILE=/tmp/.Rhistory
+
+ENV R_HOME_USER='$R_HOME'
+ENV R_LIBS_USER='$R_LIBS'
+ENV R_ENVIRON_USER='$R_ENVIRON'
+ENV R_PROFILE_USER=/opt/conda/lib/R/etc/.Rprofile
+
+RUN hictrans --help
+
+LABEL org.opencontainers.image.authors='Roberto Rossini <[email protected]>'
+LABEL org.opencontainers.image.url='https://github.com/paulsengroup/2022-mcf10a-cancer-progression'
+LABEL org.opencontainers.image.documentation='https://github.com/paulsengroup/2022-mcf10a-cancer-progression'
+LABEL org.opencontainers.image.source='https://github.com/paulsengroup/2022-mcf10a-cancer-progression'
+LABEL org.opencontainers.image.licenses='MIT'
+LABEL org.opencontainers.image.title="${CONTAINER_TITLE:-hictrans}"
+LABEL org.opencontainers.image.version="${CONTAINER_VERSION:-latest}"