SPADE: spatial deconvolution for domain specific cell-type estimation

Yingying Lu,Qin M. Chen,Lingling An

DOI: https://doi.org/10.1038/s42003-024-06172-y

IF: 6.548

2024-04-18

Communications Biology

Abstract:Understanding gene expression in different cell types within their spatial context is a key goal in genomics research. SPADE (SPAtial DEconvolution), our proposed method, addresses this by integrating spatial patterns into the analysis of cell type composition. This approach uses a combination of single-cell RNA sequencing, spatial transcriptomics, and histological data to accurately estimate the proportions of cell types in various locations. Our analyses of synthetic data have demonstrated SPADE's capability to discern cell type-specific spatial patterns effectively. When applied to real-life datasets, SPADE provides insights into cellular dynamics and the composition of tumor tissues. This enhances our comprehension of complex biological systems and aids in exploring cellular diversity. SPADE represents a significant advancement in deciphering spatial gene expression patterns, offering a powerful tool for the detailed investigation of cell types in spatial transcriptomics.

biology

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to accurately estimate the cell - type proportions within specific regions in spatial transcriptomics data. Specifically, the existing spatial transcriptomics techniques have limitations in providing cell - type - specific information, especially in the ability to capture gene expression patterns at single - cell resolution. This has led to the fact that when analyzing gene expression differences in different spatial locations, the observed changes may be affected not only by the spatial location but also by the differences in cell - type categories or proportions. Therefore, a method is required that can accurately depict and describe the spatial patterns of gene expression changes while taking into account the specificity of individual cell types.

To solve this problem, the researchers developed a new computational tool - SPADE (SPAtial DEconvolution), which accurately estimates cell - type proportions at different locations by integrating single - cell RNA sequencing (scRNA - seq), spatial transcriptomics, and histology data. The special feature of SPADE is that it utilizes an advanced spatial domain detection algorithm, combining gene expression patterns, spatial coordinates, and histology data to identify the associations between specific cell types and specific locations, the associations between spatial locations and cell types, and the similarities between adjacent locations. In addition, SPADE also includes an adaptive cell - type selection step, which can effectively determine the occurrence of specific cell types within each point.

Through this method, SPADE not only deepens the understanding of complex biological systems but also enhances the insight into the cell dynamics and composition of tumor tissues, providing a powerful tool for exploring cell diversity.