PlantscRNAdb 4.0: Improved marker identification and annotation under a cell-type uniformity for plants.

Single-cell RNA sequencing (scRNA-seq) has revolutionized plant biology by enabling high-resolution analysis of cellular heterogeneity, gene expression, and regulatory networks across diverse species. Here, we present PlantscRNAdb release 4.0, an updated comprehensive database that incorporates a new plant ontology of cell types (POCT) and covers 33 plant species. Using HCMarker, a novel computational tool developed in this study that employs a multi-metric scoring system for robust and accurate identification of marker genes across diverse plant cell types, we identified 58 846 high-confidence marker genes (with 31 818 assigned to POCT cell types) from 107 scRNA-seq datasets. We also provide cell-type-specific functional pathway analyses, revealing key biological roles and regulatory networks in different cell populations. In addition, we introduce PCmaster_anno, a deep-learning-based tool that harnesses the database's extensive data to improve automated plant cell-type annotation. Compared with previous releases, PlantscRNAdb release 4.0 not only adds 25 species and 52 datasets but also provides a more standardized and precise framework for interpreting plant cell types and their markers.