Benchmarking computational methods for detecting spatial domains and domain-specific spatially variable genes from spatial transcriptomics data.

ABSTRACT: Advances in spatially resolved transcriptomics (SRT) have led to the emergence of numerous computational methods for identifying spatial domains and spatially variable genes (SVGs); however, a comprehensive assessment of existing methods is lacking. We comprehensively benchmarked 19 methods for detecting spatial domains and domain-specific SVGs from SRT data, using 30 real-world datasets covering six SRT technologies and 27 synthetic datasets. We first evaluated the performance of these methods on spatial domain identification in terms of accuracy, stability, generalizability, and scalability. Results reveal that there is no single method that works best for all datasets, and the optimal method depends on the data, especially the SRT platform. Further, we proposed a quantitative strategy t

SUBMITTER: Kang L 

PROVIDER: S-EPMC12000868 | biostudies-literature | 2025 Apr

REPOSITORIES: biostudies-literature