Date/Time
Date(s) - 24/09/2024
3:30 pm - 4:30 pm

Location: MDCL 3020

Date/Time: Tuesday, September 24, 2024, 3.30 – 4.30 p.m.

Title: Assessing Spatial Invariance in Post-Constraint Clustering: A Framework for Robust Spatial Analysis

Speaker: Rajitha Senanayake and Pratheepa Jeganathan

Abstract: Spatial omics data provides comprehensive insights into gene or protein expression patterns across multiple tissues, which are essential for detecting tissue domains. Traditional methods, such as constraint spatial hierarchical clustering, have characterized tissue domains by classifying them in patients and associating them with clinical factors like survival time and treatment type. However, a significant limitation of these methods is their inability to uncover spatial invariant partitions. To address this limitation, we propose a robust framework for spatial analysis, which leverages constraint clustering followed by a spatial invariance test-based re-partitioning procedure.

The spatial invariance test assesses the robustness of partitions by applying multivariate spatial permutation tests to evaluate the consistency of partition results across different spatial configurations. Based on this test, clusters are re-partitioned to ensure spatial consistency, providing more reliable spatial partitions of gene or protein expressions and improving tissue domain detection.

We demonstrate the effectiveness of this framework using Multiplexed Ion Beam Imaging-Time of Flight (MIBI-TOF) data, highlighting its potential to improve tissue domain characterization. This robust approach enhances the interpretability of spatial omics data, offering a deeper understanding of the spatial distribution and interaction of gene or protein expressions within tissues.