Briefings in Bioinformatics, Год журнала: 2024, Номер 25(6)

Опубликована: Сен. 23, 2024

Abstract Single-cell RNA sequencing (scRNA-seq) technology is one of the most cost-effective and efficacious methods for revealing cellular heterogeneity diversity. Precise identification cell types essential establishing a robust foundation downstream analyses prerequisite understanding heterogeneous mechanisms. However, accuracy existing warrants improvement, highly accurate often impose stringent equipment requirements. Moreover, unsupervised learning-based approaches are constrained by need to input number prior, which limits their widespread application. In this paper, we propose novel algorithm framework named WLGG. Initially, capture underlying nonlinear information, introduce weighted distance penalty term utilizing Gaussian kernel function, maps data from low-dimensional space high-dimensional linear space. We subsequently Lasso constraint on regularized graphical model enhance its ability characteristics. Additionally, utilize Eigengap strategy predict obtain predicted labels via spectral clustering. The experimental results 14 test datasets demonstrate superior clustering WLGG over 16 alternative methods. Furthermore, analysis, including marker gene identification, pseudotime inference, functional enrichment analysis based similarity matrix algorithm, substantiates reliability offers valuable insights into biological dynamic processes regulatory

Язык: Английский