Abstract:Single-cell RNA sequencing is a transformative technology that enables us to study the heterogeneity of the tissue at the cellular level. Clustering is used as the key computational approach to group cells under the transcriptome profiles from single-cell RNA-seq data. However, accurate identification of distinct cell types is facing the challenge of high dimensionality, and it could cause uninformative clusters when clustering is directly applied on the original transcriptome. To address such challenge, an evolutionary multiobjective deep clustering (EMDC) algorithm is proposed to identify single-cell RNA-seq data in this study. First, EMDC removes redundant and irrelevant genes by applying the differential gene expression analysis to identify differentially expressed genes across biological conditions. After that, a deep autoencoder is proposed to project the high-dimensional data into different low-dimensional nonlinear embedding subspaces under different bottleneck layers. Then, the basic clustering algorithm is applied in those nonlinear embedding subspaces to generate some basic clustering results to produce the cluster ensemble. To lessen the unnecessary cost produced by those clusterings in the ensemble, the multiobjective evolutionary optimization is designed to prune the basic clustering results in the ensemble, unleashing its cell type discovery performance under three objective functions. Multiple experiments have been conducted on 30 synthetic single-cell RNA-seq datasets and six real single-cell RNA-seq datasets, which reveal that EMDC outperforms eight other clustering methods and three multiobjective optimization algorithms in cell type identification. In addition, we have conducted extensive comparisons to effectively demonstrate the impact of each component in our proposed EMDC.

DEMOC: a deep embedded multi-omics learning approach for clustering single-cell CITE-seq data

scMCs: a framework for single-cell multi-omics data integration and multiple clusterings

Multiobjective Deep Clustering and Its Applications in Single-cell RNA-seq Data

Clustering single-cell multi-omics data via graph regularized multi-view ensemble learning

A Clustering Ensemble Method for Cell Type Detection by Multiobjective Particle Optimization

Clustering of single-cell multi-omics data with a multimodal deep learning method

Clustering single cell CITE-seq data with a canonical correlation based deep learning method

Orthogonal multimodality integration and clustering in single-cell data

Clustering CITE-seq data with a canonical correlation-based deep learning method

Effective multi-modal clustering method via skip aggregation network for parallel scRNA-seq and scATAC-seq data

A Hybrid Deep Clustering Approach for Robust Cell Type Profiling Using Single-cell RNA-seq Data

BREM-SC: A Bayesian Random Effects Mixture Model for Joint Clustering Single Cell Multi-omics Data

scMNMF: a novel method for single-cell multi-omics clustering based on matrix factorization

COEM: Cross-Modal Embedding for MetaCell Identification

Clustering single-cell RNA-seq data with a model-based deep learning approach

ScMOGAE: A Graph Convolutional Autoencoder-Based Multi-omics Data Integration Framework for Single-Cell Clustering

Learning Consistency and Specificity of Cells from Single-cell Multi-omic Data

scDRMAE: integrating masked autoencoder with residual attention networks to leverage omics feature dependencies for accurate cell clustering

Clustering single-cell multi-omics data with MoClust

Deep Embedded Clustering with Multiple Objectives on Scrna-Seq Data.

Single-Cell Transcriptome Data Clustering via Multinomial Modeling and Adaptive Fuzzy K-Means Algorithm