Consensus clustering of single-cell RNA-seq data by enhancing network affinity

Abstract

Elucidation of cell subpopulations at high resolution is a key and challenging goal of single-cell ribonucleic acid (RNA) sequencing (scRNA-seq) data analysis. Although unsupervised clustering methods have been proposed for de novo identification of cell populations, their performance and robustness suffer from the high variability, low capture efficiency and high dropout rates which are characteristic of scRNA-seq experiments. Here, we present a novel unsupervised method for Single-cell Clustering by Enhancing Network Affinity (SCENA), which mainly employed three strategies: selecting multiple gene sets, enhancing local affinity among cells and clustering of consensus matrices. Large-scale validations on 13 real scRNA-seq datasets show that SCENA has high accuracy in detecting cell populations and is robust against dropout noise. When we applied SCENA to large-scale scRNA-seq data of mouse brain cells, known cell types were successfully detected, and novel cell types of interneurons were identified with differential expression of gamma-aminobutyric acid receptor subunits and transporters. SCENA is equipped with CPU + GPU (Central Processing Units + Graphics Processing Units) heterogeneous parallel computing to achieve high running speed. The high performance and running speed of SCENA combine into a new and efficient platform for biological discoveries in clustering analysis of large and diverse scRNA-seq datasets.

Detecting Fear-Memory-Related Genes from Neuronal scRNA-seq Data by Diverse Distributions and Bhattacharyya Distance

“…First 3530 Snap25+ neurons were clustered by using the SCENA method [ 71 ] and the results were plotted via UMAP [ 67 ] ( Figure 3 a). The numbers of TRAPed FR and NF cells in the five largest clusters are shown in Figure 3 b.…”

Multi-objective Bayesian Optimization with Heuristic Objectives for Biomedical and Molecular Data Analysis Workflows

“…However, there is no quantitative way to choose which parameter setting is "best" and so the community turns to a number of heuristic objectives to quantify the performance of a workflow. For example, Cui et al (7) attempt to optimize the adjusted Rand index (ARI) with respect to expert annotations and a heuristic based around downsampling rare cell types while minimizing runtime. Germain et al (1) similarly consider the ARI but also the average silhouette width to maximize cluster purity.…”

scMelody: An Enhanced Consensus-Based Clustering Model for Single-Cell Methylation Data by Reconstructing Cell-to-Cell Similarity

“…Recent advancements in ensemble clustering ( Ghaemi et al, 2009 ; Vega-Pons and Ruiz-Shulcloper, 2011 ; Boongoen and Iam-On, 2018 ) have demonstrated that integrating various basic cell partitions in a consensus matrix is effective to generate improved clustering solutions ( Kiselev et al, 2017 ; Zhu et al, 2020 ; Cui et al, 2021 ; Wang et al, 2021 ). The rationale for this idea is to construct a cell-to-cell pairwise similarity matrix based on the diverse basic clustering results through a cluster-based similarity partitioning algorithm (CSPA) ( Strehl and Ghosh, 2002 ), with each value in the matrix representing the probability of the occurrence of cell pairs in the same cluster.…”