Points2Regions: Fast, interactive clustering of imaging-based spatial transcriptomics data

Axel Andersson,Andrea Behanova,Christophe Avenel,Jonas Windhager,Filip Malmberg,Carolina Wählby

DOI: https://doi.org/10.1101/2022.12.07.519086

2024-02-15

Abstract:Imaging-based spatial transcriptomics techniques generate image data that, once processed, results in a set of spatial points with categorical labels for different mRNA species. A crucial part of analyzing downstream data involves the analysis of these point patterns. Here, biologically interesting patterns can be explored at different spatial scales. Molecular patterns on a cellular level would correspond to cell types, whereas patterns on a millimeter scale would correspond to tissue-level structures. Often, clustering methods are employed to identify and segment regions with distinct point-patterns. Traditional clustering techniques for such data are constrained by reliance on complementary data or extensive machine learning, limiting their applicability to tasks on a particular scale. This paper introduces ‘Points2Regions’, a practical tool for clustering spatial points with categorical labels. Its flexible and computationally efficient clustering approach enables pattern discovery across multiple scales, making it a powerful tool for exploratory analysis. Points2Regions has demonstrated efficient performance in various datasets, adeptly defining biologically relevant regions similar to those found by scale-specific methods. As a Python package integrated into TissUUmaps and a Napari plugin, it offers interactive clustering and visualization, significantly enhancing user experience in data exploration. In essence, Points2Regions presents a user-friendly and simple tool for exploratory analysis of spatial points with categorical labels.

Bioinformatics

What problem does this paper attempt to address?

The paper introduces Points2Regions, a practical tool for fast and interactive clustering of imaging-based spatiotemporal transcriptomic data. This data is represented as spatial points, with each point having category labels of different mRNA species. The paper aims to address the problem of analyzing these patterns of points, particularly in discovering biologically interesting patterns at different spatial scales.

Traditional clustering methods rely on additional data or complex machine learning, limiting their application in specific scale tasks. Points2Regions overcomes this challenge through its flexible and computationally efficient clustering method, which allows for pattern discovery across multiple scales and is suitable for exploratory analysis. It is designed as a Python package integrated into TissUUmaps and Napari plugins, providing interactive clustering and visualization to enhance users' data exploration experience.

The paper mentions various imaging techniques such as multiplex immunohistochemistry staining, cyclic immunofluorescence staining, imaging mass cytometry, and imaging-based spatiotemporal transcriptomics, which typically produce data that needs to be processed into point outputs and then identified into regions with different point patterns, such as cell types and tissue structures.

The effectiveness of Points2Regions is demonstrated through experiments with simulated and real data in the paper. It performs well in terms of speed and accuracy compared to existing clustering methods, without the need for extra data or complex machine learning. Additionally, it allows users to perform clustering without relying on cell segmentation, making the tool applicable in a wider range of scenarios.

In conclusion, the main objective of the paper is to propose a user-friendly, simple, and multi-scale analysis tool to explore spatial point data with category labels, in order to reveal biologically relevant regions.