Standards & Tools

Standards and tools that let researchers record cell type annotations — and the evidence behind them — in a consistent, machine-readable, reusable form.

Standards#

Cell Annotation Schema (CAS)#

A general, open LinkML schema for cell type annotations and their supporting metadata, developed as part of the scFAIR initiative to standardise single-cell genomics metadata. CAS lets researchers record not just the cell type labels assigned to cells, but the rationale and evidence behind each annotation — including marker genes used as evidence and details of automated annotation transfer. It is a compact, validatable file that links to a cell-by-gene matrix, and is structured so it can be decomposed into tabular form or flattened onto obs in AnnData. A limited-required-field core is complemented by extensions (e.g. BICAN and CAP).

Status: In development. GitHub

Tools#

CAS-tools#

A Python utility package for working with the Cell Annotation Schema, available on PyPI (pip install cas-tools) as both a library and a command-line tool. It generates CAS from CELLxGENE h5ad files and Allen Brain Cell Atlas taxonomy tables, validates annotations and their marker genes against linked AnnData, generates tabular reports from CAS, and handles interconversion between CAS and the flattened CAP-AnnData representation used by the Cell Annotation Platform.

Status: Released. GitHub

MapMyCells2CL#

A Python library and CLI that annotates MapMyCells output with Cell Ontology (CL) terms. MapMyCells assigns cells to Allen Brain Atlas taxonomy nodes; this tool maps those node IDs to CL or Provisional Cell Ontology (PCL) terms, using information-content ranking to pick the most specific applicable term. It works on both CSV/JSON MapMyCells outputs and h5ad files, adding CELLxGENE-schema-compliant cell type columns.

Status: Usable — early release (pip install mapmycells2cl). GitHub