Dockerise and conda-lock (#6)

* fix hrdetect pvalue bug

* fix run scripts

* fix cli modules

* fix README cli

* cleanup

* cleanup README

* reorg conda

* rename gha file

* add docker/conda-lock GH Action

* add Dockerfile

.Rbuildignore

@@ -12,3 +12,5 @@
 ^conda$
 ^\.github$
 ^vignettes$
+^Dockerfile$
+^\.dockerignore$

.bumpversion.cfg

@@ -9,3 +9,19 @@ replace = Version: {new_version}
 [bumpversion:file:conda/recipe/meta.yaml]
 search = version: {current_version}
 replace = version: {new_version}
+
+[bumpversion:file:.github/workflows/deploy.yaml]
+search = VERSION: '{current_version}'
+replace = VERSION: '{new_version}'
+
+[bumpversion:file:conda/env/yaml/sigrap.yaml]
+search = r-sigrap =={current_version}
+replace = r-sigrap =={new_version}
+
+[bumpversion:file:README.Rmd]
+search = sigrap.R {current_version}
+replace = sigrap.R {new_version}
+
+[bumpversion:file:README.md]
+search = sigrap.R {current_version}
+replace = sigrap.R {new_version}

.dockerignore

@@ -0,0 +1,3 @@
+*
+!conda
+!.dockerignore

.github/workflows/deploy.yaml

@@ -0,0 +1,120 @@
+name: conda-docker-docs
+
+on:
+  push:
+    branches:
+      - main
+env:
+  atoken: ${{ secrets.ANACONDA_UPLOAD_TOKEN }}
+  recipe_path: conda/recipe
+  env_yaml_path: conda/env/yaml
+  env_lock_path: conda/env/lock
+  VERSION: '0.0.4' # versioned by bump2version
+  GDRIVE_CREDENTIALS_DATA: ${{ secrets.GDRIVE_CREDENTIALS_DATA }}
+
+jobs:
+  condarise-dockerise-document:
+    # When merging to main and the commit message starts with 'Bump version:'
+    if: "startsWith(github.event.head_commit.message, 'Bump version:')"
+    name: Conda Docker Docs
+    runs-on: ubuntu-latest
+    defaults:
+      run:
+        shell: bash -l {0}
+
+    steps:
+      - name: Code checkout
+        uses: actions/checkout@v3
+      - name: Miniconda setup
+        uses: conda-incubator/setup-miniconda@v2
+        with:
+          auto-update-conda: true
+          add-pip-as-python-dependency: false # conda-lock goes crazy if pip pre-installed
+          activate-environment: env1
+          mamba-version: "*"
+          channels: "umccr,conda-forge,bioconda"
+      - name: 🐍 Conda pkg build and upload
+        run: |
+          conda activate
+          mamba create --name cbuildenv conda-build conda-verify anaconda-client boa dvc dvc-gdrive
+          conda activate cbuildenv
+          dvc pull
+          ls -l inst/extdata/*
+          conda mambabuild --R 4.1 ${recipe_path} --token ${atoken}
+      - name: 🔒 Conda lock
+        run: |
+          conda activate
+          mamba create --name lockenv conda-lock pip
+          conda activate lockenv
+          conda-lock --file ${env_yaml_path}/sigrap.yaml --platform linux-64
+          mv conda-lock.yml ${env_lock_path}/conda-lock.yml
+      - name: 💾 Commit lockfile
+        run: |
+          # TODO: the version bumping guarantees there will be changes,
+          # but during testing this can be problematic - need to
+          # commit only if changes exist.
+          git status
+          git config --local user.email "[email protected]"
+          git config --local user.name "GitHub Actions"
+          MSG="[bot] Updating conda-lock file (v${VERSION})"
+          git add .
+          git commit -m "${MSG}"
+          git push
+
+      - name: 🌐 Website publish
+        run: |
+          conda activate
+          mamba env create -n pkgdownenv -f ${env_yaml_path}/pkgdown.yaml
+          conda activate pkgdownenv
+          sigrap.R --version
+          Rscript -e "pkgdown::deploy_to_branch(pkg = '.', commit_message = paste(pkgdown:::construct_commit_message('.'), '- see https://umccr.github.io/sigrap/'), branch = 'gh-pages', new_process = FALSE)"
+
+      - name: 📥 Pull lock commit
+        id: pull_lock_commit
+        run: |
+          git pull --no-rebase
+          git --no-pager log --decorate=short --pretty=oneline -n3
+          LATEST_COMMIT_HASH="$(git --no-pager log -1 --format='%H')"
+          echo "::set-output name=latest_commit_hash::${LATEST_COMMIT_HASH}"
+      - name: 🔖 Create tag
+        uses: actions/github-script@v5
+        env:
+          LATEST_COMMIT_HASH: ${{ steps.pull_lock_commit.outputs.latest_commit_hash }}
+        with:
+          script: |
+            const the_tag_name = 'refs/tags/v' + process.env.VERSION
+            const the_sha = process.env.LATEST_COMMIT_HASH
+            github.rest.git.createRef({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              ref: the_tag_name,
+              sha: the_sha
+            })
+
+      # work with tag from above
+      - name: Code checkout
+        uses: actions/checkout@v3
+        with:
+          ref: v${{ env.VERSION }}
+      - name: 🏰 QEMU setup
+        uses: docker/setup-qemu-action@v2
+      - name: 🏯 Buildx setup
+        uses: docker/setup-buildx-action@v2
+        with:
+          install: true
+          config-inline: |
+            [worker.oci]
+              max-parallelism = 2
+      - name: 🐙 GitHub CR login
+        uses: docker/login-action@v2
+        with:
+          registry: ghcr.io
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+      - name: 🐳 Docker img build and push
+        uses: docker/build-push-action@v2
+        with:
+          tags: ghcr.io/${{ github.repository }}:${{ env.VERSION }}
+          context: . # yes, dot
+          push: true
+          platforms: linux/amd64

.gitignore

@@ -4,3 +4,4 @@
 .Ruserdata
 /docs
 /nogit
+README.html

Dockerfile

@@ -0,0 +1,32 @@
+FROM ubuntu:20.04
+LABEL maintainer="https://github.com/pdiakumis"
+
+ARG MINI_VERSION=4.12.0-0
+ARG MINI_URL=https://github.com/conda-forge/miniforge/releases/download/${MINI_VERSION}/Mambaforge-${MINI_VERSION}-Linux-x86_64.sh
+ARG MAMBA_PREFIX="/opt/mambaforge"
+
+# install core pkgs, mambaforge
+RUN apt-get update -qq && \
+    apt-get install --yes --no-install-recommends -qq \
+    git bash bzip2 curl ca-certificates && \
+    apt-get clean && \
+    rm -r /var/lib/apt/lists/* && \
+    rm -r /var/cache/apt/* && \
+    curl --silent -L "${MINI_URL}" -o "mambaforge.sh" && \
+    /bin/bash mambaforge.sh -b -p "${MAMBA_PREFIX}" && \
+    rm mambaforge.sh
+
+# create conda env
+ENV PATH="${MAMBA_PREFIX}/bin:$PATH"
+ARG ENV_DIR=/home/sigrap_conda_env
+ARG ENV_NAME="sigrap_env"
+COPY ./conda/env/lock ${ENV_DIR}
+RUN conda config --set always_yes yes && \
+    mamba install -c conda-forge conda-lock==1.0.5
+RUN conda-lock install --name ${ENV_NAME} ${ENV_DIR}/conda-lock.yml && \
+    mamba clean --all --force-pkgs-dirs
+
+ENV PATH="${MAMBA_PREFIX}/envs/${ENV_NAME}/bin:${PATH}"
+ENV CONDA_PREFIX="${MAMBA_PREFIX}/envs/${ENV_NAME}"
+
+CMD [ "sigrap.R" ]

R/hrdetect.R

@@ -147,7 +147,7 @@ hrdetect_read_purple_cnv <- function(x) {
 #' (l <- hrdetect_prep_snvindel(x, nm = "sampleA"))
 #' @testexamples
 #' expect_equal(c("snv_results", "indel_results"), names(l))
-#' expect_equal(c("sig", "exposure", "pvalue"), colnames(l[["snv_results"]]))
+#' expect_equal(c("sig", "exposure"), colnames(l[["snv_results"]]))
 #' expect_equal(colnames(l[["indel_results"]])[c(1, 7)], c("sample", "del.mh.prop"))
 #'
 #' @export
@@ -169,10 +169,12 @@ hrdetect_prep_snvindel <- function(x, nm = NULL, genome = "hg38",
     subs = snvindel_tabs[["snv"]],
     genome.v = genome
   )[["catalogue"]]
-  assertthat::assert_that(inherits(snv_catalogue, "data.frame"),
-                          ncol(snv_catalogue) == 1,
-                          colnames(snv_catalogue) == "catalogue",
-                          nrow(snv_catalogue) == 96)
+  assertthat::assert_that(
+    inherits(snv_catalogue, "data.frame"),
+    ncol(snv_catalogue) == 1,
+    colnames(snv_catalogue) == "catalogue",
+    nrow(snv_catalogue) == 96
+  )
 
   subs_fit_res <- signature.tools.lib::Fit(
     catalogues = snv_catalogue,
@@ -184,7 +186,8 @@ hrdetect_prep_snvindel <- function(x, nm = NULL, genome = "hg38",
 
   assertthat::assert_that(
     length(subs_fit_res) == 16,
-    "exposures" %in% names(subs_fit_res))
+    "exposures" %in% names(subs_fit_res)
+  )
 
   snv_exp <- subs_fit_res[["exposures"]] |>
     t() |>
@@ -305,8 +308,10 @@ hrdetect_run <- function(nm, snvindel_vcf, sv_vcf, cnv_tsv, genome = "hg38",
     genome <- "hg19"
   }
 
-  snvindel <- hrdetect_prep_snvindel(x = snvindel_vcf, nm = nm,
-                                     genome = genome, sigsToUse = sigsToUse)
+  snvindel <- hrdetect_prep_snvindel(
+    x = snvindel_vcf, nm = nm,
+    genome = genome, sigsToUse = sigsToUse
+  )
   snv <- snvindel[["snv_results"]] |>
     dplyr::filter(.data$sig %in% c("Signature3", "Signature8")) |>
     tidyr::pivot_wider(

R/mutationalpatterns.R

@@ -292,7 +292,7 @@ sig_workflow_run <- function(vcf, sample_nm, ref_genome = "hg38", outdir) {
     }
   }
 
-  cli::cli_h2(glue::glue("{date_log()} Start of MutationalPatterns workflow"))
+  cli::cli_h2(glue::glue("{gpgr::date_log()} Start of MutationalPatterns workflow"))
   # import VCF
   gr <- MutationalPatterns::read_vcfs_as_granges(
     vcf_files = vcf,
@@ -355,7 +355,7 @@ sig_workflow_run <- function(vcf, sample_nm, ref_genome = "hg38", outdir) {
     }() |>
     sigrap::sig_contribution_table(type = "ID")
 
-  cli::cli_h2(glue::glue("{date_log()} Saving MutationalPatterns results to\n'{outdir}'"))
+  cli::cli_h2(glue::glue("{gpgr::date_log()} Saving MutationalPatterns results to\n'{outdir}'"))
   save_plot_list(p_snv, file.path(outdir, "plot/snv"))
   save_plot_list(p_dbs, file.path(outdir, "plot/dbs"))
   save_plot_list(p_indel, file.path(outdir, "plot/indel"))
@@ -364,5 +364,5 @@ sig_workflow_run <- function(vcf, sample_nm, ref_genome = "hg38", outdir) {
   gpgr::write_jsongz(x = sigs_dbs, path = file.path(outdir, "sigs/dbs.json.gz"))
   gpgr::write_jsongz(x = sigs_indel, path = file.path(outdir, "sigs/indel.json.gz"))
 
-  cli::cli_h2(glue::glue("{date_log()} End of MutationalPatterns workflow"))
+  cli::cli_h2(glue::glue("{gpgr::date_log()} End of MutationalPatterns workflow"))
 }