C3: An R package for cross-species compendium-based cell-type identification

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Compendium-based cell type identification relies on gene expression databases, which are scarce for non-model organisms.
• Existing methods face challenges in transcriptome-based sample classification for species lacking extensive data.

Purpose of the Study:

• To address the challenge of cell type identification in non-model organisms with limited data.
• To introduce a cross-species compendium comparison strategy using the C3 R package.

Main Methods:

• Utilized a cross-species gene set analysis (XGSA) framework to correct for homologous gene mapping biases.
• Implemented the framework as an open-source R package named C3.
• Evaluated C3 performance using public data from the NCBI Gene Expression Omnibus.

Main Results:

• C3 demonstrated effectiveness in cell type identification across different species.
• Comparative analysis showed C3 performs comparably or better than existing gene expression profile matching tools.
• The package provides a simple and effective method for transcriptome-based sample classification.

Conclusions:

• C3 offers a robust solution for cell type identification in non-model organisms by enabling cross-species analysis.
• The open-source R package facilitates broader research applications in comparative genomics.
• Cross-species compendium comparison is a viable strategy to overcome data limitations in under-resourced species.