This repository contains a pipeline that inputs a variants.csv file and applies the following steps:

1. Split it so there won't be over 10k variants (FABIAN limitation).

2. Run it through FABIAN, against the following TF's: AR CBX3 DMRT1 DMRT3 ESR1 ESR2 FOXL1 GATA4 LHX9 NR5A1 RUNX1 SOX10 SOX8 SOX9 SRY WT1

3. Process the results into the original variants file.

4. Produce an excel file with a heatmap, marking the most significant scores.

The output file is the same variants.csv file with extra columns, one for each TF, that contain the average gain/loss of function score + "fabian_heatmap.xlsx".

Run with the command:

nextflow run fabian_analysis.nf --csvFile path/to/variants.csv --curProcessedOutputDir results_dir