Cell composition analysis of bulk genomics using single-cell data.

Area of Science:

• Computational Biology
• Immunology
• Genomics

Background:

• Single-cell RNA sequencing (scRNA-seq) provides deep insights into cellular heterogeneity in complex tissues.
• Understanding cell-type composition is crucial for studying tissue responses to stimuli like viral infections.

Purpose of the Study:

• To introduce Cell Population Mapping (CPM), a novel deconvolution algorithm.
• To leverage scRNA-seq reference profiles for inferring cell composition from bulk transcriptome data.
• To analyze cell-state dynamics and their relationship with clinical outcomes during influenza infection.

Main Methods:

• Development of the Cell Population Mapping (CPM) deconvolution algorithm.
• Application of CPM using the 'scBio' CRAN R-package.
• Analysis of bulk transcriptome data from influenza-virus-infected mouse lungs.
• Mathematical modeling to explain observed cell-state dynamics and clinical outcomes.

Main Results:

• CPM successfully infers cell type and state composition from bulk transcriptome data.
• A cell-state-specific, gradual variation was observed in the relationship between cell abundance and clinical symptoms in infected mouse lungs.
• This gradual change was validated experimentally and explained by a mathematical model linking cell-state dynamics to clinical outcomes.

Conclusions:

• CPM is a powerful tool for reconstructing the continuous spectrum of cell states in heterogeneous tissues.
• The study highlights the importance of considering cell-state dynamics, not just cell abundance, in understanding disease.
• Findings provide a framework for analyzing complex tissue responses using bulk transcriptomics and scRNA-seq references.