Ranking of cell clusters in a single-cell RNA-sequencing analysis framework using prior knowledge

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Ranking cell types in single-cell RNA sequencing (scRNA-seq) is crucial for biological discovery but existing methods have limitations.
• Traditional approaches include analyzing cell type proportions, differentially expressed genes, or relying solely on prior knowledge, which can be insufficient.

Purpose of the Study:

• To develop and present a novel methodology for prioritizing and ranking cell types identified through scRNA-seq analysis.
• To integrate prior biological knowledge, including molecular mechanisms and drug information, with experimental data for improved cell type relevance.
• To enhance the identification of cell types most pertinent to specific disease contexts.

Main Methods:

• Developed a new methodology that combines prior knowledge (molecular mechanisms, drugs) with expert user input for scRNA-seq data analysis.
• Ranked cell types based on the alignment of their expression profiles with disease-associated molecular mechanisms and drugs.
• Incorporated cell-cell communication network perturbations between disease and control states to further refine cell type prioritization.

Main Results:

• The novel methodology effectively ranks cell types by relating their expression profiles to disease-specific prior knowledge.
• The approach allows for rapid and automated prioritization of biologically meaningful cell types.
• Cell-cell communication analysis further aids in the ranking and prioritization of cell types.

Conclusions:

• The presented methodology offers a significant advancement over traditional cell ranking techniques in scRNA-seq analysis.
• It provides a complementary approach that leverages prior knowledge in a rapid and automated fashion.
• The methodology is implemented as the R package scRANK, available on GitHub, facilitating its application in disease research.