To create the Conda environment, run the following command:

conda env create -f env.yaml
conda activate my_metadensity2

pre-commit install

Pre-commit is a framework for managing and maintaining Git pre-commit hooks. It automatically runs checks (like code formatting, linting, or other validations) on your code before you commit it to the repository. This ensures code quality and consistency across the project.

• lint your jupyter notebook inline by %load_ext lab_black

• follow this instruction to install git-lfs.
• and configure

• scripts: contains snakemake commands to run large scale analysis including kicking off Skipper, overlap analysis etc
• configs: Skipper configs
• tables: intermediate tables

• scripts: Contains Snakemake commands for running large-scale analyses, including initiating Skipper, overlap analysis, etc.
• configs: Configuration files for Skipper.
• tables: Intermediate tables generated during analysis.

• 0_RBP_annotation: Counting eCLIPs, annotating GO/PPI/domains, aligning IDs, and processing other datasets.
• 1_eCLIP_QC: Quality control of eCLIP data.
• 2_eCLIP_clusters: Clustering and t-SNE plots (Figure 1).
• 2_eCLIP_motifs: Motif analysis using HOMER, SELEX/RBNS, and RBPNet seqlets.
• 2_eCLIP_overlap: Overlapping binding sites across eCLIP peaks.
• 2_RBP_mutation: gnomAD analysis on RBPs, including o/e ratio calculations (Figure 2).
• 3_RNAseq: Legacy folder (RNAseq is not included in this paper).
• 4_CLIP_ML_RBPNet: Training, evaluation, model comparison, and benchmarking (Figures 3 and 4).
• 5_gnomAD_reference_scaling: Scaling o/e and MAPs ratios to reproduce past gnomAD results (not a main finding).
• 6_gnomAD_popgen_analysis: Selection analysis in binding sites (Figure 5).
• 7_ClinVar_analysis: Pathogenic variant analysis in binding sites (Figure 6).