the current workflow for running SCENIC is based on this github repo from the aerts lab

It uses nextflow to execute time-intensive processing followed by downstream analysis using a modifiable jupyter notebook. Input is a loom file, possibly converted from a seurat object with cells of interest. output of the nextflow pipeline is another loom file.

A helper script, run_scenic.sh is included to ease execution of the nextflow pipeline and standardize file paths for input and output loom files. run_scenic.sh also includes scenic-required reference files located at ~/Homo_sapiens/scenic/ that can be swapped out if needed.

Quick Start

1. navigate to project directory
2. clone this repository into the src subdirectory git clone https://github.com/whtns/scenic_src.git
3. run run_scenic.sh with a command-line argument indicating the path to the desired seurat object bash run_scenic.sh unfiltered
4. open the jupyter notebook scenic_downstream_analysis.ipynb in jupyter lab located on the srt server at cobrinik-srt.la.ad.chla.org:8888 and modify as needed to explore common regulons. modify paths as needed to load `output/scenic/-final.loom