credtools chunk

Find independent loci, split summary statistics by locus, and optionally extract LD matrices.

credtools chunk INPUT_CONFIG OUTPUT_DIR [OPTIONS]

Common Use

credtools chunk work/munged/sumstat_info_updated.txt work/chunks --threads 4

Inputs

Use a population config with:

popu    cohort  sample_size path    ld_ref

path should point to munged summary statistics. ld_ref should be a PLINK prefix if you want LD extraction.

Options

Option	Meaning	Default
--distance	distance threshold for independent loci, in bp	500000
--pvalue	p-value threshold for lead variants	5e-8
--merge-overlapping	merge overlapping loci across inputs	on
--use-most-sig	use the best SNP if no SNP passes the threshold	on
--min-variants	minimum variants required in a locus	10
--custom-chunks	TSV with chr, start, end	none
--ld-format	LD reference format, usually plink	plink
--keep-intermediate	keep temporary LD extraction files	off
--threads	worker count	1
--log-file	write logs to a file	none

Outputs

OUTPUT_DIR/
- identified_loci.txt
- loci_list.txt
- chunks/
- prepared/

loci_list.txt is the handoff file for pipeline, meta, qc, and finemap.

Custom Regions

Use custom chunks when you already know the regions:

chr start   end
1   50000000    50500000
9   21900000    22100000

credtools chunk population_config.tsv work/chunks --custom-chunks regions.tsv