Convert GWAS summary statistics to the variant call format

A common format to store GWAS summary data and associated metadata to aid with data sharing and tool development (manuscript in preparation)

The VCF format has a number of benefits:

• Pre-existing, well known and well defined format
• Handles complex variants, build differences and multi-allelics using the HTSLIB library
• Indexing makes querying by position extremely fast: VCF is 8.6-45.5x faster to extract variant(s) by genomic position
• Many established tools already exist for manipulation
• The file format is relatively small
• Indexing time is very fast
• Multiple GWAS may be stored in a single file
• Easy to export data into other tabular formats
• Can import directly into a distributed database: GenomicsDB

Requirements

This tool expects GWAS summary statistics in tabular format, one variant per row. The minimum required fields are:

• Chromosome
• Base position (currently only hg19/GRCh37 is supported)
• Allele
• Effect size
• Standard error
• P value

Usage

Users may run the tool on their own hardware by downloading Gwas2VCF

For online submissions:

1. Fill in the form below, provide as much information as possible
2. Record the unique job identifier
3. Check job status
4. Download GWAS-VCF
5. Parse GWAS-VCF in R or Python and connect to downstream analyses using gwasglue

Example

1. Download a subset of the GLGC (Willer CJ et al, 2013) GWAS of LDL cholesterol:

curl -o 'example.txt' 'https://raw.githubusercontent.com/MRCIEU/gwas2vcfweb/master/app/tests/data/example.1k.txt'

2. Complete submission fields or click populate
3. Select file for upload
4. Submit form

Populate