Neuroelectrophysiology Analysis Ontology (NEAO) for data sharing and reproducibility

• The Neuroelectrophysiology Analysis Ontology (NEAO) targets neuroelectrophysiology data sharing and reproducibility.¹
• The ontology enables interoperable pipelines and reuse across labs and tooling.¹
• Nature-published; clear implementation path for electrophysiology data and analysis.^{1 1}

Weekly enrichment (2026-07-20)

• The peer-reviewed paper appeared in Nature’s Scientific Data (Köhler, Grün, Denker; Forschungszentrum Jülich / INM-6), with an earlier open preprint on arXiv (2412.05021).^{2 3 4}
• NEAO is a set of OWL controlled vocabularies published under the persistent namespace http://purl.org/neao, organized into modular submodules for analysis steps, data entities, parameters, and bibliographic references imported by a top-level neao.owl file.^{2 5}
• The ontology’s stated goal is unambiguous, machine-readable description of neuroelectrophysiology analysis independent of the specific software, function names, or programming language, addressing the problem that different codes implement the same algorithm under different parameter names.^{2 4}
• NEAO explicitly targets the FAIR principles—improving findability, interoperability, and reuse of analysis results—by classifying processes so results can be compared across labs and toolchains.^{2 4}
• It is designed to consume automatically captured provenance: the Alpaca (Automatic Lightweight Provenance Capture) framework records atomic Python analysis steps and serializes them with the W3C PROV standard, which NEAO then semantically enriches.⁶
• Real-world demonstrations annotate complex workflows built on the Elephant electrophysiology library and the Cobrawap cortical-wave analysis pipeline, showing knowledge-graph querying over heterogeneous results.^{2 6}
• Source development is centralized in the public GitHub repository INM-6/neuroephys_analysis_ontology, with XML catalog files enabling correct local imports in Protégé.⁵
• BCI/neuroinformatics implication: standardized, queryable provenance for spike/LFP/ECoG analysis eases reproducibility and cross-study comparison for neural-interface pipelines; the authors note the approach could extend to other computational workflows such as neural simulation.^{2 6}

Footnotes

1. https://news.google.com/rss/articles/CBMiX0FVX3lxTFBjYl9LLS1pek5sRGdvc0x3cFFTbThDX1ZEZUpybG9obTNFS0V5X2pGR2N2dVRJTTNrLXduZ2tMWXYyYXp5c2EwTUZIRy1jT3VLUkQ2S0lTWUpHMmpYWXdF?oc=5 ↩ ↩² ↩³ ↩⁴

2. https://www.nature.com/articles/s41597-025-05213-3 ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

3. https://doi.org/10.1038/s41597-025-05213-3 ↩

4. https://arxiv.org/html/2412.05021v1 ↩ ↩² ↩³

5. https://github.com/INM-6/neuroephys_analysis_ontology ↩ ↩²

6. https://doi.org/10.5281/zenodo.16736245 ↩ ↩² ↩³