File Formats

Annotation Files

remeta uses many of the same files as regenie to define variant sets. Most annotation files compatible with regenie should also be compatible with remeta.

--anno-file

1:55039839:T:C PCSK9 LoF
1:55039842:G:A PCSK9 missense
.

A file defining variant annotations. Contains 3 whitespace delimited columns: variant id (in CPRA format), gene name, and variant annotation.

--set-list

A1BG 19  58346922  19:58346922:C:A,19:58346924:G:A,...
A1CF 10  50806630  10:50806630:A:G,10:50806630:A:AT,...
.

A file defining which variants are part of a gene set. Contains 4 whitespace delimited columns: gene name, chromosome, start position, and a comma separated list of variants in the gene.

--mask-def

LoF LoF
missense missense
LoF_missense LoF,missense
.

A file specifying which annotation categories to combine into masks. Contains 2 whitespace delimited columns: mask name, and a comma separated list of variant annotations in a mask.

--aaf-file

An optional file with variant alternate allele frequencies. If specified, these frequencies are used for building masks. Three formats are supported. All formats assume whitespace separated columns.

• Two Column Format: CPRA and allele frequency
• Five Column Format: CPRA, frequency, is singleton (either 0 or 1), chromosome, position
• Six Column Format (plink2 afreq file): chromosome, CPRA, ref, alt, alt_freqs, and obs_ct

Here is an example of the five column format:

7:6187101:C:T 1.53918207864341e-05 0 7 6187101
7:6190395:C:A 2.19920388819247e-06 1 7 6190395
.

Reference LD Files

--ld-prefix

A set of three files named $PREFIX.remeta.gene.ld, $PREFIX.remeta.buffer.ld, and $PREFIX.remeta.ld.idx.gz generated by remeta compute-ref-ld. The index $PREFIX.remeta.ld.idx.gz is bgzipped and human readable. The columns are:

• GENE_NAME
• GENE_LD_INDEX
• BUFFER_LD_INDEX
• GENE_VARIANTS_STORED
• BUFFER_VARIANTS_STORED

--gene-list

PCSK9   1   55039446    55064852
USP24   1   55066358    55215364
.

A file listing gene start and end positions. Contains 4 whitespace separated columns: gene name, chromosome, start position, end position. Note that this file must align with the --set-list file for gene-based tests. Specifically,

1. Gene names in the --gene-list file should match gene names in the --set-list file exactly.
2. Any variant in the --set-list must appear within the start position and end position of the LD matrix for that gene: otherwise it will be dropped from the test (unless the --keep-variants-not-in-ld-mat option is specified).
3. The start and end position should be inclusive of any variant that could appear in a test. In particular, it is not recommended to set the start and end positions based on the variants in the set list. If the set list changes, this could result in variants being dropped from a test.

In the remeta repository we provide an example gene list under resources/Ensembl100.GRCh38.chr1_23.gene_list.txt.gz directory. This file was created by extracting gene boundaries for protein coding genes annotated in Ensembl release 100. An Ensembl GTF file was downloaded from Ensembl; start and end positions were extracted from lines with feature gene and the biotype protein_coding.

--target-pfile and --buffer-pfile

A set of pgen, pvar, and psam files from plink2.

--genetic-map

A genetic map in the SHAPEIT format. Contains 3 columns: position, chromosome, and centimorgan. Note that genetic maps are available from the SHAPEIT5 repository.

Miscellaneous Files

--htp

A file in htp format, the default output format of remeta. Output by regenie with the --htp option. htp files are whitespace separated with the following columns:

• Name: Variant name. For remeta, single variants must be encoded CHROM:POS:REF:ALT, matching the naming convention of the annotation files.
• Chr: Chromosome.
• Pos: Position.
• Ref: Reference allele.
• Alt: Alternate allele.
• Cohort: Name of the cohort (e.g. UK Biobank).
• Model: Association model.
• Effect: Effect size. Note that for binary traits these are odds-ratios, not log odds ratios.
• LCI_effect: Lower 95% confidence interval for the effect size.
• UCI_effect: Upper 95% confidence interval for the effect size.
• Pval: Association p-value.
• AAF: Alternate allele frequency.
• Num_Cases: Number of cases. For QTs this is the sample size.
• Cases_Ref: Number of cases homozygous for the reference allele.
• Cases_Het: Number of cases who are heterozygotes.
• Cases_Alt: Number of cases homozygous for the alternate allele.
• Num_Controls: Number of controls. For QTs this is NA (as are the fields below).
• Controls_Ref: Number of controls homozygous for the reference allele.
• Controls_Het: Number of controls who are heterozygotes.
• Controls_Alt: Number of controls who are homozygous for the alternate allele.
• Info: A semi-colon (;) separated list of arbitrary fields. These can be KEY=VALUE pairs or a single field alone.

--extract and --exclude

Files with variant ids (one per line) to include or exclude from meta-analysis.

--genep-def

LoF LoF
NonSyn LoF,missense,LoF_missense
.

A file defining which masks to combine with ACAT. Contains two space-separated columns: the name of the GENEP set, and a comma separated list of masks to include.