Snakemake RNA modification regression pipeline

Snakemake pipeline for performing regression in RNA modification data

Quickstart

Quick demo shown below.

1. Set up conda environments shown in envs.

# Starting from this directory
SNK_REPO=$(pwd)

cd envs

# Option 1: direct conda
bash environment_setup.sh

# Option 2: docker (use if running on mac M*)
docker build -t rna_mod_gsis --platform linux/amd64 .

2. Run demo

# Starting from this directory
SNK_REPO=$(pwd)

# set up the directory with the required data dir
cd demo
ln -s data/demo_data/demo_data-random1k.h5ad data/data.h5ad

# Run analysis
bash run_gwas__local_docker.sh

Working directory structure

• Required input files: in the data directory.
• Final output data: in the results directory.
• Logs of all jobs: in the logs directory.

Input

AnnData .h5 file with the following attributes:

• Layers contains those specified under left hand side of formula configuration. Example: counts_modified and counts_unmodified in the demo dataset
• Observation slot (.obs) contains columns specified on right hand side of formula configuration. Example: glucose_condition, cell_type, and library_size in the demo dataset
• X slot contains total counts of the two test layers. Example: X contains counts_modified + counts_unmodified in the demo dataset