output	github_document

barb-mars

Repo of Collaboration with Barbieri. The pipeline use cellranger v7 to align the fastq.

Project uses {targets}and {renv} and {here}.

data/ contains the reference and a user provided csv mapping that indicates the location of the samples, as well as the results of alignment.

• mapping_folders.csv and mapping_folders_2.csv are used in one of the targets to map folders containing fastqs to specific sample names.
• batch_1_2_aggr.csv is used by the target that runs cellranger aggr
• reference and refdata_* contains the references used for alignment.
• **alignment_results/ **contains the outputs of cellranger alignment. Each folder is the alignment of one sample. The samples are then aggregated together in the alignment_results/batch_1_2 folder.

code/ functions called by _targets.R in here::here("code", "targets_functions"))

Howto

The pipeline uses {targets} (https://books.ropensci.org/targets/) and is setup to run in parallel on SLURM (with {future.batchtools}).

The files batchtools.slurm.tmpl and .batchtools.conf.R are configuration files used by the API {future.batchtools} to communicate with SLURM. Insert path to your home directory in .batchtools.conf.R and copy batchtools.slurm.tmpl to you home directory.

Resources to deploy for each target are specified in the file _targets_resources.conf.R that is sourced in the main _targets.R file.

Use targets::tar_make_future(workers = 10) from inside srun interactive session or run sbatch start_targets.sh to initiate the pipeline. The job master will control the execution of the other jobs.

SLURM Logs are created in log/ (be sure to have a folder with that name in the current directory). A reproducible R environment is maintained using {renv} Initialize and install packages with:

renv::restore()

Run the pipeline.

targets::tar_make()

Check manifest with targets::tar_manifest() and Load targets with targets::tar_load("name_of_target)