SifiNet: A robust and accurate method to identify feature gene sets and annotate cells.

Area of Science:

• Computational biology
• Genomics
• Bioinformatics

Background:

• Accurate identification of cellular subpopulations is crucial for understanding complex biological systems.
• Existing computational pipelines often rely on cell clustering, which can introduce inaccuracies.
• There is a need for robust methods that can analyze multiomic single-cell data effectively.

Approach:

• SifiNet is a computational pipeline designed for gene set identification, cell subpopulation annotation, and relationship elucidation.
• It uniquely bypasses the cell clustering stage to avoid potential inaccuracies.
• The pipeline supports both single-cell RNA sequencing (scRNA-seq) and single-cell ATAC sequencing (scATAC-seq) data.

Key Points:

• SifiNet demonstrates superior performance compared to state-of-the-art methods across multiple experimental datasets.
• It enables comprehensive multiomic cellular profiling by integrating scRNA-seq and scATAC-seq data.
• The method provides robust identification of distinct gene sets and cellular subpopulations.

Conclusions:

• SifiNet offers a more accurate and robust approach to single-cell data analysis by omitting the clustering step.
• The pipeline facilitates deeper insights into cellular heterogeneity and inter-population relationships.
• As an open-source R package, SifiNet is accessible for broad application in biological research.