Cell type discovery using single-cell transcriptomics: implications for ontological representation.

Area of Science:

• Genomics
• Cell Biology
• Bioinformatics

Background:

• Multicellular organisms rely on diverse cell types for physiological functions.
• Single-cell RNA sequencing (scRNA-seq) generates large datasets for novel human cell type discovery.
• Understanding cell types is crucial for human health and disease research.

Purpose of the Study:

• To review recent advancements in human cell type characterization using scRNA-seq and single-nuclei RNA sequencing (snRNA-seq).
• To discuss the impact of these discoveries on the Cell Ontology (CL).
• To propose a machine learning-based method for defining cell types.

Main Methods:

• Review of recent literature on scRNA-seq and snRNA-seq in human systems.
• Analysis of implications for the Cell Ontology (CL).
• Development of a random forest machine learning approach for marker gene identification.

Main Results:

• scRNA-seq and snRNA-seq are rapidly identifying novel human cell types.
• These discoveries necessitate updates to the Cell Ontology (CL).
• A machine learning method can identify marker genes for consistent cell type definitions.

Conclusions:

• Standardized cell type definitions are essential for integrating high-throughput biological data.
• The proposed method facilitates the creation of FAIR (findable, accessible, interoperable, reusable) cell type classes.
• The Cell Ontology (CL) can serve as a central knowledgebase for cellular phenotypes in health and disease.