Pathway-Based Single-Cell RNA-Seq Classification, Clustering, and Construction of Gene-Gene Interactions Networks...

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-Seq) provides high-resolution gene expression data for cellular analysis.
• Existing machine learning methods primarily focus on clustering cells, with limited exploration of functional pathways for heterogeneity classification.
• Understanding cellular heterogeneity is crucial for interpreting diverse biological functions.

Purpose of the Study:

• To develop a pathway-based analytical framework for scRNA-Seq data using machine learning.
• To identify discriminative functional pathways and gene-gene interactions (GGIs) related to cellular heterogeneity.
• To cluster cell populations and pinpoint pivotal genes driving cellular functions.

Main Methods:

• Utilized Random Forests (RF) for a pathway-based analysis of scRNA-Seq data.
• Developed a novel method to construct GGIs networks using RF to illustrate differentiating interactions.
• Identified co-occurring genes in discriminative pathways and 'cross-talk' genes connecting them.

Main Results:

• The proposed framework effectively clusters cell populations based on pathway information.
• Identified key functional pathways and pivotal genes contributing to cellular heterogeneity.
• Generated networks illustrating GGIs, co-functional genes, and cross-talk between pathways.

Conclusions:

• The pathway-based framework enhances the understanding of functional heterogeneity in scRNA-Seq data.
• It provides a robust method for prioritizing pathways and identifying key genes in cellular differentiation.
• This approach facilitates deeper biological interpretation of complex cell populations.