Output Files

CREDTOOLS writes a lot of files because each step keeps its handoff files. This page explains the files you will actually open most often.

After munge

work/munged/
- EUR_UKBB.munged.txt.gz
- AFR_MVP.munged.txt.gz
- sumstat_info_updated.txt

File	Use it for
*.munged.txt.gz	cleaned summary statistics
sumstat_info_updated.txt	input to credtools chunk

After chunk

work/chunks/
- identified_loci.txt
- loci_list.txt
- sumstat_info_updated.txt
- chunks/
- prepared/

File or directory	Use it for
identified_loci.txt	see which regions were found
chunks/	locus-sized summary statistics
prepared/	prepared summary statistics, LD matrices, and LD maps
loci_list.txt	input to pipeline, meta, qc, or finemap

If ld_ref was present, chunk also tries to build LD files.

After prepare

work/prepared/
- prepared_files.txt
- loci_list.txt
- EUR.locus_1.sumstats.gz
- EUR.locus_1.ld.npz
- EUR.locus_1.ldmap.gz

File	Use it for
prepared_files.txt	status table for prepared locus file sets
loci_list.txt	input to pipeline, meta, qc, or finemap
*.sumstats.gz	locus-level munged summary statistics
*.ld.npz	compressed LD matrix
*.ldmap.gz	LD variant map and reference allele frequencies

After meta

work/meta/
- loci_info.txt
- heterogeneity.txt.gz
- {locus_id}/
  - {prefix}.sumstats.gz
  - {prefix}.ld.npz
  - {prefix}.ldmap.gz

File	Use it for
loci_info.txt	input to QC or fine-mapping after meta-analysis
heterogeneity.txt.gz	cross-study heterogeneity summary
{locus_id}/...	meta-analyzed locus files

After qc

work/qc/
- qc.txt.gz
- qc_run_summary.log
- {locus_id}/
- cleaned/

Common files:

File	Use it for
qc.txt.gz	one summary table across loci
{locus_id}/qc.txt.gz	per-locus QC summary
{locus_id}/expected_z.txt.gz	kriging RSS LD and z-score consistency
{locus_id}/dentist_s.txt.gz	Dentist-S outlier statistics
{locus_id}/compare_maf.txt.gz	allele-frequency checks
cleaned/cleaned_loci_info.txt.gz	input after --remove-outlier

The cleaned/ directory appears only when outlier removal is requested and outliers are found.

After pipeline

work/results/
- overall_run_summary.log
- {locus_id}/
  - pips.txt.gz
  - credible_sets_summary.txt.gz
  - causal_variants.txt.gz
  - parameters.json
  - run_summary.log
  - expected_z.txt.gz
  - dentist_s.txt.gz
  - compare_maf.txt.gz

Start with these:

File	First question it answers
overall_run_summary.log	did all loci finish?
{locus_id}/run_summary.log	did this locus finish?
{locus_id}/pips.txt.gz	which variants have high PIP?
{locus_id}/credible_sets_summary.txt.gz	how many credible sets were found?
{locus_id}/causal_variants.txt.gz	which variants are in credible sets?
{locus_id}/parameters.json	which tool and settings were used?

Reading PIP Tables

pips.txt.gz includes SNPID, PIP, CRED, and available summary-statistic columns. For multi-input results, study-specific columns are prefixed by population and cohort.

gzip -cd work/results/locus_1/pips.txt.gz | head

CRED = 0 means the variant is not assigned to a credible set. CRED = 1 means it belongs to the first credible set.

Reading Credible Set Summaries

gzip -cd work/results/locus_1/credible_sets_summary.txt.gz | head

Use this file for reports. Use pips.txt.gz when you need variant-level detail.