scAce: an adaptive embedding and clustering method for single-cell gene expression data.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for cell type identification.
• Existing clustering methods often require pre-defined cluster numbers or initial assignments, limiting their flexibility.

Purpose of the Study:

• To develop an adaptive embedding and clustering method for scRNA-seq data.
• To overcome limitations of existing methods by removing the need for predetermined cluster numbers.

Main Methods:

• Proposed scAce, a variational autoencoder-based method for simultaneous cell embedding and clustering.
• Developed an adaptive cluster merging approach for improved clustering without prior number estimation.
• Implemented an optional clustering enhancement feature to refine assignments.

Main Results:

• scAce demonstrated superior performance compared to state-of-the-art methods on simulated and real scRNA-seq datasets.
• Achieved higher clustering accuracy and robustness in cell type identification.
• Successfully identified cell types without requiring the number of clusters to be specified beforehand.

Conclusions:

• scAce offers an effective and robust solution for scRNA-seq data clustering.
• The adaptive nature of scAce enhances its applicability and performance in cell type discovery.
• The method provides a valuable tool for the analysis of single-cell gene expression data.