About ONGA

ONGA (Ontology for Genomic Annotations) provides controlled vocabularies for describing genomic data files. It defines two complementary vocabularies: - DataType: Computational output types (HOW data was produced) - FeatureType: Biological feature types (WHAT biology it represents) A file's DataType describes its computational origin (e.g., "peaks" from a peak caller). A file's FeatureType describes what biological phenomenon it captures (e.g., "TF binding sites"). The same DataType can represent different FeatureTypes depending on the experiment.

The Problem

When you have a genomic data file (like a BED file), what does it actually contain? Is it peaks from ChIP-seq? ATAC-seq? IDR-filtered peaks? Methylation states? Signal tracks?

The ENCODE consortium developed approximately 310 internal terms for describing file contents ("output types"), but these terms aren't formalized as an ontology, lack definitions, and aren't easily usable outside ENCODE.

Existing ontologies like EDAM cover broad bioinformatics concepts but lack the granularity needed for real-world genomic file annotation.

The Solution

ONGA formalizes this vocabulary with:

Scope

In Scope

Out of Scope (use existing ontologies)

Technical Details

Version
0.1.0
License
CC BY 4.0
Namespace
https://databio.org/onga/
Format
LinkML
Created by
Nathan Sheffield
Created
2026-05-20

Links

Contact

For questions, suggestions, or to propose new terms, please open an issue on GitHub.