GWAS Gold Standards

WORK IN PROGRESS

GWAS Gold Standards contained in the repository

We assembled a repository of >400 published GWAS loci for which we have high confidence in the gene functionally implicated. Gold-standard evidence were grouped into 4 classes: (i) GWAS loci that overlap with drug targets-disease pairs; (ii) expert curated loci with strong orthogonal evidence or biological plausibility; (iii) GWAS loci that have been investigated by functional follow up experiments (e.g. Reporter Assays, CRISPR/Cas9 genome editing); (iv) loci inferred from observational functional data (e.g. colocalisation with molecular QTLs and epigenetics marks). We also assigned each gold-standard a confidence rating of high, medium or low depending on our assessment of the strength of supporting evidence.

Repository for GWAS variant to gene gold standards.

gold_standards/processed contains the latest set of gold standards in a variety of formats. Note, the TSV file has arrays concatenated together using '|' as a seporator.

To do:

Scripts to produce descriptive statistics for each gold standard set
Derive new gold standards based on drug and rare disease data

Gold standard schema

The gold standard schema consists of 5 sections:

sentinel_variant: Information about the sentinel/lead GWAS variant
trait_info: Information about the GWAS trait or disease
association_info: Evidence linking the sentinel variant to the trait through GWAS
gold_standard_info: Evidence linking the GWAS signal (variant, trait, association) to a gene
metadata: Additional information

Example schema in yaml format

# Sentinel variant information, alleles must match gnomAD
sentinel_variant:
  # Locus chrom and position on either GRCh37 or GRCh38 are required
  locus_GRCh37:
    chromosome: '16'
    position: 81264597
  locus_GRCh38:
    chromosome: '16'
    position: 81230992
  # Alleles are required
  alleles:
    alternative: G
    reference: T
  # rsID is optional
  rsid: rs6564851

# Trait/disease information
trait_info:
  # List of ontology codes for the trait, should be EFO if possible
  ontology:
  - HMDB0000561
  # Trait reported by the author and standardised trait name (from ontology)
  reported_trait_name: Carotenoid and tocopherol levels (beta-carotene)
  standard_trait_name: B-Carotene

# Association evidence
association_info:
  # List of ancestries in which the association was detected
  ancestry:
  - EUR
  # GWAS Catalog study ID if available
  gwas_catalog_id: GCST000324
  # Open Targets Genetics study ID if available
  otg_id: GCST000324
  # Negative log p-value (optional)
  neg_log_pval: 23.699
  # Pubmed ID or doi
  pubmed_id: '19185284'
  doi: '10.1101/592238'

# Gold standard evidence
gold_standard_info:
  # Ensembl gene ID
  gene_id: ENSG00000135697
  # List of evidences support link with gene
  evidence:
  # Item 1 in list
  - class: expert curated # See below for evidence classes
    # Confidence should be "High" or "Low"
    confidence: High
    # Evidence curator
    curated_by: EF
    # Description of evidence
    description: BCO1 (previously referred to as BCMO1) encodes beta-carotene oxygenase
      1 which uses a molecule of oxygen to produce two molecules of retinol from
      beta-carotene.  Enzyme deficiency results in accumulation of beta-carotene.
    # Pubmed ID or source
    pubmed_id: '11401432'
    source: ChEMBL drug data

# Metadata
metadata:
  date_added: '2019-05-17'
  reviewed_by: EM
  # Name given to the group of gold standards
  set_label: ProGeM
  submitted_by: EF
  # Additional tags that may be useful for analysis
  tags:
  - metabolite
  - mQTL
  comments: 'No comments'

Possible evidence classes

"expert curated": association curated by an expert
"functional experimental": association inferred from experimental alteration (intervention), e.g. CRISPR editing
"functional observational": association inferred from observational evidence, e.g. correlation with quantitative trait such as eQTL or pQTL
"drug": association inferred from known drug target-indication pairs

How to submit a new gold standard

Download the template yaml from here if submitting a single gold standard or here if submitting multiple at once
Fill in the yaml file
Create a new issue including the completed yaml file

We will then review the submission and add it to the repository.

Validate and process new gold standards

The sections contains instructions for validating and processing newly submitted gold standards. Submitters are not required to validate and process new gold stnadards.

# Set up environment
conda env create -n goldstandards --file environment.yaml
conda activate gold_standards

# Validate against schema (input can be json or yaml)
python validation/validator.py \
  --input temp/progem/progem.190517.yaml \
  --schema validation/goldstandard_schema.v1.4.json

# Convert to json
python utils/json_yaml_converter.py \
  --input temp/progem/progem.190517.yaml \
  --output temp/progem/progem.190517.json

# Add to `gold_standards/unprocessed_validated`
mv temp/progem/progem.190517.json \
  gold_standards/unprocessed_validated/progem.190517.json

# Process all gold standards in `gold_standards/unprocessed_validated`
nano processing/process_and_convert_formats.sh # Edit Args
bash processing/process_and_convert_formats.sh

opentargets/genetics-gold-standards

GWAS Gold Standards