Shuwei Yin, Yi Lu, Wenyu Yan, Zonghui Zhu, Ruiqi Liu, GuangLin Li
Developed scLncR, an integrated and flexible framework for lncRNA analysis in single-cell and single-nucleus transcriptomic data, incorporating modules for lncRNA prediction, expression matrix processing, cell-type specific analysis, expression enrichment, weighted gene co-expression network analysis, pseudotime trajectory, and functional enrichment, available via command-line or a Shiny-based graphical user interface. Benchmarking revealed that the independent lncRNA matrix processing strategy enhances lncRNA signal visibility while preserving high concordance with established preprocessing methods at both cell-type and cluster levels. Application to Arabidopsis root datasets prioritized three lncRNA candidates linked to root-hair cellular states and distinct genetic contexts.
BACKGROUND AND AIMS: Long non-coding RNAs (lncRNAs) are important regulators of cellular processes, but their analysis at single-cell resolution remains challenging because lncRNA prediction, quantification, cell-type-specific characterization and downstream functional interpretation are often performed using separate tools. Although single-cell RNA sequencing (scRNA-seq) and single-nucleus RNA sequencing (snRNA-seq) provide cellular-resolution transcriptomic profiles, reproducible workflows for lncRNA-focused analysis, particularly in plant systems, remain limited. To address this need, we developed scLncR, an open-source, modular and reproducible framework for lncRNA analysis in single-cell and single-nucleus transcriptomic data.
METHODS: scLncR is a versatile framework incorporating multiple functional modules: lncRNA prediction, independent expression matrix processing, cell-type specific expression analysis, snRNA-seq/scRNA-seq expression enrichment analysis, weighted gene co-expression network analysis (WGCNA), pseudotime trajectory, and functional enrichment. It supports both command-line operation (for server-based customization) and a Shiny-based graphical user interface for user-friendly access.
RESULTS: By connecting discrete analytical steps, scLncR enables a seamless transition from candidate lncRNA discovery to biological interpretation. Benchmarking revealed that our independent lncRNA matrix processing strategy enhances lncRNA signal visibility while preserving high concordance with established preprocessing methods at both cell-type and cluster levels. Notably, application to Arabidopsis root datasets prioritized three lncRNA candidates linked to root-hair cellular states and distinct genetic contexts.
CONCLUSIONS: scLncR serves as an open-source workflow resource designed to systematize and streamline lncRNA-focused analyses for single-cell and single-nucleus transcriptomic data. The source code, configuration files and documentation available at https://github.com/Lilab-SNNU/scLncR, release v1.0.0.