PoweREST: Statistical power estimation for spatial transcriptomics experiments to detect differentially expressed genes between two conditions

Recent advancements in spatial transcriptomics (ST) have significantly enhanced biological research in various domains. However, the high cost for current ST data generation techniques restricts the large-scale application of ST. Consequently, maximization of the use of available resources to achieve robust statistical power for ST data is a pressing need. One fundamental question in ST analysis is detection of differentially expressed genes (DEGs) under different conditions using ST data. Such DEG analyses are performed frequently, but their power calculations are rarely discussed in the literature. To address this gap, we developed PoweREST, a power estimation tool designed to support the power calculation for DEG detection with 10X Genomics Visium data. PoweREST enables power estimation both before any ST experiments and after preliminary data are collected, making it suitable for a wide variety of power analyses in ST studies. We also provide a user-friendly, program-free web application that allows users to interactively calculate and visualize study power along with relevant parameters.Author summary: Spatial transcriptomics technologies provide an unprecedented view of gene expression in tissues while preserving spatial context, enabling important discoveries in various biomedical fields, especially cancer research. However, the cost of profiling a single spatial transcriptomics slice typically ranges from $7,500 to $14,000, highlighting the importance of careful experimental design during the early planning stages. Over-sampling can lead to unnecessary financial waste, while under-sampling risks insufficient statistical power, potentially resulting in a failure to detect true biological information. To address this challenge, we introduce a computational framework that estimates the statistical power for detecting differentially expressed genes in spatial transcriptomics experiments. Our method accounts for key factors when planning spatial transcriptomics studies, such as spatial information of gene expression within regions of interest, log-fold changes in gene expression between experimental conditions, gene detection rates, and number of slice replicates. In addition to a software package, we also provide a user-friendly, program-free web application that allows users to interactively calculate and visualize study power.