A layered standards framework for integrating single-cell and spatial omics data into brain cell atlases.

Area of Science:

• Neuroscience
• Bioinformatics
• Data Science

Background:

• The BRAIN Initiative Cell Atlas Network (BICAN) generates large-scale multimodal brain cell data.
• Data integration is challenging due to diverse assays, experimental contexts, and distributed infrastructure.

Purpose of the Study:

• To implement a standards framework for integrating and reusing BICAN data.
• To create interoperable brain cell atlases from heterogeneous datasets.

Main Methods:

• Developed an assay-agnostic modeling layer for omics data representation.
• Implemented harmonized metadata standards for biospecimen provenance.
• Created an extensible representation for data-driven cell type taxonomies.

Main Results:

• Established an end-to-end framework transforming heterogeneous data into structured resources.
• Enabled data reuse via mapping algorithms, annotation systems, and visualization platforms.
• Linked biospecimen provenance with cell-level outputs in a standardized taxonomy.

Conclusions:

• The framework facilitates an efficient data-to-knowledge pipeline for large-scale consortia.
• This approach supports broad community reuse and knowledge-driven analysis of brain cell data.
• Demonstrated a generalizable strategy for data integration and atlas creation.