Ó³»­´«Ã½

Abstract:

Advances in high-throughput spatial transcriptomics (ST) technologies enable high-throughput molecular profiling of cells while maintaining their spatial organization within tissues. Such high-throughput ST data demand new statistical approaches and computational analysis methods to identify genes that spatially change in their expression patterns between conditions, such as in diseased versus healthy tissues. In this talk, I will provide an overview of the latest ST computational analysis methods developed by my lab. In particular, to facilitate spatial molecular comparisons across structurally matched tissue sections from replicates, case-control settings, and within and across technologies, we previously develop STalign to align ST datasets in a manner that accounts for partially matched tissue sections and other local non-linear distortions using diffeomorphic metric mapping. Likewise, to enhance the scalability of ST data analysis, we developed a rasterization preprocessing framework called SEraster that aggregates cellular information into spatial pixels. More recently, we developed STcompare to integrate STalign and SEraster into a statistical framework for comparative analysis of ST data by testing for differences in spatial correlation and spatial fold-change across structurally matched locations while robustly controls for false positives even in the presence of spatial autocorrelation common in ST data. Alternatively, to facilitate spatial molecular comparisons across structurally unmatched tissues, we previously developed CRAWDAD, Cell-type Relationship Analysis Workflow Done Across Distances, to quantify cell-type spatial relationships across multiple length scales. We have applied CRAWDAD to compare cell-type spatial organizations across samples as well as across functional tissue units within samples. Overall, we anticipate that such computational methods for analyzing ST data will contribute to important biological insights regarding spatial molecular changes across comparative axes of interest.