Single-cell RNA-seq clustering: datasets, models, and algorithms.

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) offers unprecedented opportunities for biological discovery.
• Analyzing scRNA-seq data is crucial for understanding cell heterogeneity, identifying subgroups, and dynamics.
• Clustering is a key analytical step for scRNA-seq data interpretation.

Purpose of the Study:

• To review relevant datasets and analytical tools for scRNA-seq data analysis.
• To discuss various scRNA-seq clustering models, including K-means, hierarchical, and consensus clustering.
• To compare the performance of state-of-the-art scRNA clustering methods.

Main Methods:

• Review of popular scRNA-seq datasets.
• Discussion of scRNA-seq clustering algorithms.
• Comparative analysis of seven scRNA clustering methods on five public datasets.
• Evaluation using Adjusted Rand Index (ARI) and Normalized Mutual Information (NMI).

Main Results:

• Seven advanced scRNA clustering methods were evaluated.
• Performance metrics (ARI and NMI) were used for comparison.
• Unsupervised models effectively cluster scRNA-seq data but present challenges.

Conclusions:

• Clustering is vital for scRNA-seq data analysis, revealing cell heterogeneity.
• Comparative analysis provides insights into the strengths and weaknesses of different clustering methods.
• Future research should address existing challenges in scRNA-seq clustering.