rmogire/liver-enzyme-gwas

Template scripts for “Multi-cohort GWAS of liver-enzyme traits in African-ancestry populations.” Reproducible, well-commented SAIGE→METASOFT→SuSiE→FastENLOC→LDSC workflow; no datasets included. # Liver-Enzyme GWAS (African-Ancestry) – Template Scripts _Companion code for Mogire RM *et al.*_ This repository contains **fully documented templates** that reproduce every analysis step we ran for the manuscript **“Multi-cohort genome-wide association analyses reveal loci underlying circulating liver enzyme levels in African-ancestry populations.”** > ⚠️ **No data are included.** > Each script has placeholder paths (`path/to/...`). > Replace those with your own file locations before running. --- ## Directory map | Path | Purpose | |--------------------------|-------------------------------------------| | `scripts/` | Bash / Python / R templates for each step | | `configs/` | Optional config files (e.g. FUMA) | | `results/ logs/` | Created at run-time; ignored by Git | --- ## Template overview | Stage | Script / file | |------------------------------------|---------------| | Single-cohort GWAS (SAIGE) | `scripts/template_run_gwas_saige.sh` | | Meta-analysis (METASOFT) | `scripts/template_meta_analysis_metasoft.sh` | | Fine-mapping (SuSiE + PLINK) | `scripts/template_finemap_susie.py` | | eQTL colocalisation (FastENLOC) | `scripts/template_colocalization_fastenloc.sh` | | SNP-heritability & rg (LDSC) | `scripts/template_ldsc_rg.sh` or `template_ldsc_workflow.sh` | | Manhattan / QQ plots (R) | `scripts/template_plot_manhattan_qq.R` | Each file starts with a **usage block** showing the exact command to run after you have edited the paths. --- ## Software / environment A minimal conda specification is provided in `environment.yml` with: * Python ≥ 3.10 (pandas, numpy, seaborn, matplotlib, gwaslab, rpy2) * R ≥ 4.2 (susieR, data.table, qqman) * SAIGE, METASOFT, PLINK 2, LDSC * Java 8 (for METASOFT) > **FastENLOC** must be compiled separately; see its GitHub page. Create and activate …