Merge pull request #953 from jashapiro/jashapiro/seurat-convert

Add Seurat conversion script

.github/workflows/docker_seurat-conversion.yml

@@ -13,22 +13,22 @@ concurrency:
   cancel-in-progress: true
 
 on:
-  # pull_request:
-  #   branches:
-  #     - main
-  #   paths:
-  #     - "analyses/seurat-conversion/Dockerfile"
-  #     - "analyses/seurat-conversion/.dockerignore"
-  #     - "analyses/seurat-conversion/renv.lock"
-  #     - "analyses/seurat-conversion/conda-lock.yml"
-  # push:
-  #   branches:
-  #     - main
-  #   paths:
-  #     - "analyses/seurat-conversion/Dockerfile"
-  #     - "analyses/seurat-conversion/.dockerignore"
-  #     - "analyses/seurat-conversion/renv.lock"
-  #     - "analyses/seurat-conversion/conda-lock.yml"
+  pull_request:
+    branches:
+      - main
+    paths:
+      - "analyses/seurat-conversion/Dockerfile"
+      - "analyses/seurat-conversion/.dockerignore"
+      - "analyses/seurat-conversion/renv.lock"
+      - "analyses/seurat-conversion/conda-lock.yml"
+  push:
+    branches:
+      - main
+    paths:
+      - "analyses/seurat-conversion/Dockerfile"
+      - "analyses/seurat-conversion/.dockerignore"
+      - "analyses/seurat-conversion/renv.lock"
+      - "analyses/seurat-conversion/conda-lock.yml"
   workflow_dispatch:
     inputs:
       push-ecr:

.github/workflows/run_seurat-conversion.yml

@@ -19,15 +19,15 @@ concurrency:
 
 on:
   workflow_dispatch:
-  # workflow_call:
-  # pull_request:
-  #   branches:
-  #     - main
-  #   paths:
-  #     - analyses/seurat-conversion/**
-  #     - "!analyses/seurat-conversion/Dockerfile"
-  #     - "!analyses/seurat-conversion/.dockerignore"
-  #     - .github/workflows/run_seurat-conversion.yml
+  workflow_call:
+  pull_request:
+    branches:
+      - main
+    paths:
+      - analyses/seurat-conversion/**
+      - "!analyses/seurat-conversion/Dockerfile"
+      - "!analyses/seurat-conversion/.dockerignore"
+      - .github/workflows/run_seurat-conversion.yml
 
 jobs:
   run-module:
@@ -44,19 +44,27 @@ jobs:
           r-version: 4.4.0
           use-public-rspm: true
 
-      - name: Set up pandoc
-        uses: r-lib/actions/setup-pandoc@v2
+      - name: Install additional dependencies
+        run: |
+          sudo apt-get install -y \
+            libhdf5-dev \
+            libglpk40
 
       - name: Set up renv
         uses: r-lib/actions/setup-renv@v2
         with:
           working-directory: ${{ env.MODULE_PATH }}
 
-      # Update this step as needed to download the desired data
       - name: Download test data
-        run: ./download-data.py --test-data --format SCE
+        env:
+          #  Unless otherwise stated, the first sample ID in the given project
+          #  For the multiplexed project SCPCP000009, set of samples `SCPCS000133,SCPCS000134,SCPCS000135,SCPCS000136` (this group of samples has 2 associated libraries). Only `SCPCS000133` is specified in the SAMPLES string here.
+          #   For the project SCPCP000011, the sample SCPCS000435 was specifically selected as an edge case sample with <10 cells
+          SAMPLES: "SCPCS000001,SCPCS000024,SCPCS000050,SCPCS000101,SCPCS000124,SCPCS000133,SCPCS000168,SCPCS000216,SCPCS000246,SCPCS000250,SCPCS000298,SCPCS000435,SCPCS000481,SCPCS000484,SCPCS000490,SCPCS000502,SCPCS000514,SCPCS000616,SCPCS000632,SCPCS000758"
+        run: |
+          ./download-data.py --test-data --format SCE --samples ${SAMPLES}
 
       - name: Run analysis module
         run: |
           cd ${MODULE_PATH}
-          # run module script(s) here
+          bash run_seurat-conversion.sh

analyses/seurat-conversion/.Rprofile

@@ -2,3 +2,8 @@
 if (Sys.getenv('OPENSCPCA_DOCKER') != 'TRUE') {
   source('renv/activate.R')
 }
+
+# work around renv 1.0.11 bug
+if (exists("object")){
+  rm(object)
+}

analyses/seurat-conversion/README.md

@@ -4,29 +4,27 @@ This analysis module is used to convert the SingleCellExperiment objects from th
 
 ## Description
 
-Please provide a description of your module, including:
 
-- What type of analysis is included?
-- What methods or tools are used?
-
-If there are multiple steps in the module, please include an outline of the analysis steps and scripts used.
 
 ## Usage
 
-TBD
+The `run_seurat-conversion.sh` script will convert all `_processed.rds` and `_filtered.rds` files in the repository's `data/current/` directory, placing output files in the module's `results/seurat/` directory.
+To process specific sets of files, the `scripts/seurat-conversion.R` script can be run directly, specifying input and output directories as needed.
 
-## Input files
 
-Input modules will be a SingleCellExperiment objects (as `.rds` files) from the ScPCA Portal.
+## Input files
 
+Input files are SingleCellExperiment objects (as `.rds` files) from the ScPCA Portal.
+By default, these are the files in the `data/current/` directory of the `OpenScPCA-analysis` repository, as downloaded by the `download-data.py` script.
 
 ## Output files
 
-Seurat object files (as `.rds` files) will be generated as output, organized by project.
+Seurat object files (as `.rds` files) will be generated as output, organized following the same directory structure as the input files.
+For example, if an input file is `$REPO_ROOT/data/current/SCPCP000000/SCPCS000000/SCPCL000000_processed.rds`, the corresponding output file will be placed at `results/seurat/SCPCP000000/SCPCS000000/SCPCL000000_processed_seurat.rds`.
 
 ## Software requirements
 
-The code for the project will be in R, so all software dependencies will be managed by `renv`.
+The code for this module is all written in R, with all software dependencies managed by `renv`.
 
 ## Computational resources
 

analyses/seurat-conversion/components/dependencies.R

@@ -1,2 +1,4 @@
 # R dependencies not captured by `renv`
-# library("missing_package")
+library("Seurat")
+library("scater")
+library("scran")