ScSpTITH: a rank-correlation framework for robust quantification of multi-dimensional tumor heterogeneity

Intra- and inter-tumoral heterogeneity (ITH) is a fundamental hallmark of cancer, driving spatial, temporal, cellular, and tumor microenvironmental (TME) complexity and critically contributing to therapeutic resistance. Single-cell RNA sequencing (scRNA-seq) provides unprecedented resolution for dissecting tumor heterogeneity; however, its accuracy is severely compromised by pervasive "dropout" artifacts, resulting in zero-inflation rates of 70-95% in typical scRNA-seq datasets. To address this limitation, we introduce ScSpTITH (Single-cell and Spatial Transcriptomic Intra-/inter-Tumoral Heterogeneity), a robust computational framework for quantifying ITH in both scRNA-seq and spatial transcriptomic data. ScSpTITH first selects highly variable genes based on standard deviation to prioritize biologically informative features, and then computes pairwise Spearman rank correlations across cells. This rank-based strategy confers inherent robustness to technical noise, non-normality, and high dropout rates, enabling stable and interpretable quantification of transcriptional heterogeneity. Across diverse cancer and developmental datasets, elevated ScSpTITH scores are strongly associated with active tumor evolution, advanced disease stage, pronounced cellular plasticity, therapeutic resistance, and poor clinical outcomes. ScSpTITH is scalable and flexible, allowing heterogeneity to be quantified across inter-tumoral and intra-tumoral dimensions, including spatial regions, cell populations, and treatment conditions. Collectively, ScSpTITH provides a unified and robust framework for dissecting multi-dimensional heterogeneity in single-cell and spatial transcriptomic studies.