Skip to main content
Version: next

PEG Metadata Preparation

💡 To make data submission easier, we provide an Excel template with six tabs (one per entity type), specifically designed to help submitters capture the relevant fields efficiently.

tip

📊 Download metadata google sheet template

If you have questions about any attribute for each entity, we also provide detailed explanations here. We are more than happy to hear from you — please feel free to contact us if you have further questions.

The PEG metadata is currently organised into six tabs. The diagram below shows a simplified view of their relationships.

Each entity contains fields that capture a different aspect of the dataset:

  • Dataset description - descriptors for the whole PEG matrix (trait, source of the matrix itself, publication reference, release date, creator)
  • Genomic Identifier – details about the variants, genes, or locus included in your dataset.
  • Evidence – supporting data types and experimental or computational evidence that link variants to genes or traits.
  • Integration – information about how different streams of evidence are combined (e.g., scoring, weighting, prioritisation).
  • Source – citation and provenance information for each evidence stream, including publications, databases, and biosample details.
  • Method – a description of the methodology, pipelines, or softwares used to generate the data.

How to fill the metadata template​

Below is a practical, step-by-step guide to help you complete the Excel/Google Sheet metadata template. The fields are explained in detail on the Metadata Standard page, but the guidance here focuses on what to fill and when.

1) Prepare your inputs​

Before you start filling the sheet, make sure you have:

  • The final PEG evidence matrix you plan to submit (all columns finalized)
  • The GWAS source and trait details
  • A list of evidence column names
  • A list of integration column names (if any), including which evidence they combine
  • A list of sources (databases, files, publications) and methods (software or manual pipelines) used to generate evidence/integration

2) Fill tabs in a logical order​

We recommend filling the tabs in the following order so that your source_tag and method_tag to avoid repeats:

  1. Dataset description
  2. Genomic Identifier
  3. Evidence
  4. Integration
  5. Source
  6. Method

Tab-by-tab guidance​

Dataset description (required)​

Describe the trait, GWAS source, and overall PEG dataset context.

  • trait_description is mandatory and should be clear to a non-specialist.
  • gwas_source should be a GWAS Catalog accession (GCST) if possible.
  • If gwas_source is not GCST, fill in gwas_sample_description, gwas_sample_size, and gwas_sample_ancestry.

Genomic Identifier (required)​

Explain how variants, genes, and loci were defined.

  • genome_build is mandatory (e.g., GRCh38).
  • Include variant_type (lead/index/sentinel/mixed), and how loci are defined.
  • If you use a specific gene ID version (Ensembl, HGNC), record it here.

Evidence (required)​

This is the core tab. Each row represents one evidence column in your PEG matrix.

  • column_header must exactly match the evidence column name in your matrix.
  • column_description explain the data content and how to interperate the data
  • evidence_category, evidence_category_abbreviation must come from the controlled list.
  • variant_or_gene_centric is important for Other_[CustomisedCategory]_(stream)_[details].
  • evidence_stream_tag identifies the specific analysis stream within an evidence category. It repeats the stream defined in the Category_(stream)_[details] column header.
  • source_tag and method_tag are optional but highly recommended, as they provide additional detail in the Source and Method tabs and help others better understand and reuse your data.
  • Add threshold and note if the values require interpretation.

Integration (required if you have integrated results)​

Each row describes one integration output column in the PEG matrix.

  • integration_tag is a short identifier for the integration method.
  • column_header should follow the format INT_[integration_tag]_[details].
  • method_tag is mandatory and must point to the method used for integration.
  • Use evidence_streams_included (pipe-delimited) to show which streams were combined.
  • Use integrations_included (pipe-delimited) to show which integrated analysis also included.
  • Only one row in the integration tab can have author_conclusion = True (this defines the PEG list).

Record the origin of each evidence stream. One row = one source definition.

  • source_tag must be unique and is referenced in Evidence/Integration tabs.
  • If multiple files belong to the same source, duplicate the row and update only file_name / url.
  • Use the biosample section if tissue/cell type is relevant to that source.

Describe how each evidence stream or integration was produced.

  • method_tag must be unique and referenced later.
  • method_mode and method_mode_ontology_term_id are required (They are prepared by the drop down).
  • If method_mode = software, include name/version/URL.
  • If method_mode = manual, describe the process in method_description.

Final checks before submission​

  • All column_header values match the evidence/integration columns in your PEG matrix.
  • source_tag and method_tag values are unique and used consistently.
  • Evidence categories and abbreviations match the controlled list.
  • Only one integration column is marked author_conclusion = True.
  • All mandatory fields are filled.

If you are unsure about any field, refer back to the PEG Metadata Standard or contact us.