Computational approaches for the analyses of spatial profiling technologies

Division Of Basic Sciences - Nci

Investigators

Linked publications, trials & patents

Abstract

The recent progress in multiplexed protein imaging has advanced our ability to identify topological structures in tissue microenvironments associated with distinct clinical characteristics. A significant challenge with high-dimensional imaging is how to systematically infer spatial features underlying different patient groups, like immunotherapy response vs resistance groups. We develop an artificial intelligence framework for identifying differential protein patterns between distinct groups from multiplexed proteomics images. Our image region-based framework does not require any prior manual/semi-automatic steps like cell segmentation and cell type annotation as required in existing spatial data analysis workflows and therefore can capture essential features not defined by humans. The framework is also suitable for use with a low number of labeled samples, as is the case for early-phase exploratory spatial studies. We applied the framework to identify differential protein patterns between different outcome or treatment groups in humans and mice. On a pre-treatment multiplexed imaging dataset of triple-negative breast cancer patients treated with anti-PDL1 therapy, the framework shows that a higher relative abundance of a signature region enriched in Ki67, TCF7, CD8, CD45, GATA3, MHCI, and PDL1 is significantly associated with patient pathological complete response. We expect that our proposed AI framework will be a useful tool for generating novel hypotheses regarding biomarkers or regulators of cancer immunotherapy outcomes.