A comprehensive human embryo reference tool using single-cell RNA-sequencing data

Area of Science:

• Developmental Biology
• Stem Cell Biology
• Genomics

Background:

• Stem cell-derived human embryo models are valuable tools for studying early development.
• Assessing the fidelity of these models to in vivo development is crucial.
• Single-cell RNA sequencing (scRNA-seq) is a key technology for this assessment.

Purpose of the Study:

• To develop a comprehensive, integrated human scRNA-seq reference dataset for early development.
• To establish a standardized resource for benchmarking and authenticating human embryo models.
• To create a predictive tool for annotating cell identities in novel datasets.

Main Methods:

• Integration of six published human scRNA-seq datasets (zygote to gastrula).
• Validation of lineage annotations using human and nonhuman primate data.
• Development of a prediction tool using Uniform Manifold Approximation and Projection (UMAP).

Main Results:

• A unified human scRNA-seq reference dataset spanning early embryogenesis was established.
• A predictive tool was created, enabling projection and annotation of query datasets.
• Analysis revealed risks of misannotation in existing embryo models lacking proper benchmarking.

Conclusions:

• The developed reference dataset and prediction tool are essential for accurate authentication of human embryo models.
• Utilizing this resource mitigates misannotation risks and enhances the reliability of developmental studies.
• This work provides a universal benchmark for evaluating stem cell-based human embryo models.