credtools 0.9.2

multi-ancestry fine-mapping pipeline.

CREDTOOLS

CREDTOOLS is a command-line and Python toolkit for GWAS fine-mapping workflows. It helps you standardize summary statistics, define loci, prepare LD-backed locus files, run QC, combine cohorts, and run fine-mapping tools from one repeatable interface.

Documentation: https://Jianhua-Wang.github.io/credtools

What It Does

• Standardizes raw GWAS summary statistics with credtools munge.
• Finds and chunks loci from genome-wide summary statistics with credtools chunk.
• Prepares locus-level summary statistics, LD matrices, and LD maps.
• Runs meta-analysis with meta_all, meta_by_population, or no_meta.
• Runs QC for LD and summary-statistic consistency.
• Runs fine-mapping with SuSiE, FINEMAP, ABF, CARMA, SuSiEx, MultiSuSiE, MESuSiE, and RSparsePro wrappers.
• Writes PIP tables, credible set summaries, causal-variant tables, logs, and plots.

Install

python -m pip install credtools
credtools --version

CREDTOOLS supports Python >=3.9,<3.13.

Some workflows need external tools:

Workflow	Extra dependency
LD extraction in chunk	PLINK
--tool finemap	FINEMAP executable
--tool susiex	SuSiEx executable
--tool susie_ash or susie_inf	Rscript and susieR >= 0.16.1
--tool carma	Rscript and CARMA
--tool mesusie	Rscript and MESuSiE

See the external tools guide for setup checks.

Quick Start

Start from a population config:

popu	cohort	sample_size	path	ld_ref
EUR	UKBB	400000	/data/EUR.sumstats.gz	/ref/EUR
AFR	MVP	90000	/data/AFR.sumstats.gz	/ref/AFR

Run the workflow:

credtools munge population_config.tsv work/munged
credtools chunk work/munged/sumstat_info_updated.txt work/chunks
credtools pipeline work/chunks/loci_list.txt work/results \
  --tool susie \
  --meta-method meta_all \
  --max-causal 5

Before a full run, inspect work/chunks/loci_list.txt. If chunk wrote a placeholder sample_size or cohort label, edit it before fine-mapping.

Main Commands

Command	Use it for
credtools munge	clean and standardize summary statistics
credtools chunk	find loci, create chunks, and prepare LD-backed files
credtools prepare	build LD-backed locus files from chunked sumstats and genotype references
credtools meta	combine cohorts before QC or fine-mapping
credtools qc	check summary-statistic and LD consistency
credtools finemap	run fine-mapping only
credtools pipeline	run meta-analysis, QC, and fine-mapping together
credtools plot	create QC and fine-mapping plots

Key Input

Most downstream commands use a loci_list.txt:

locus_id	chr	start	end	popu	cohort	sample_size	prefix
locus_1	1	50000000	50500000	EUR	UKBB	400000	data/EUR_UKBB_locus_1

Each prefix should point to:

{prefix}.sumstats.gz
{prefix}.ld.npz
{prefix}.ldmap.gz

See the file schema reference for exact columns and accepted file names.

Development

git clone https://github.com/Jianhua-Wang/credtools.git
cd credtools
uv sync
uv run pytest
uv run mkdocs serve

Links

• Documentation: https://Jianhua-Wang.github.io/credtools
• GitHub: https://github.com/Jianhua-Wang/credtools
• PyPI: https://pypi.org/project/credtools/
• License: MIT