A Novel Open-Source CellProfiler Pipeline for Automated, User-Friendly Hierarchical and K-Means Clustering of Microglial Morphology
Microglia represent a highly dynamic and heterogeneous cell type that is critically implicated in states of health and pathology. Microglial morphological subgroups have been identified that correspond to functional characteristics determining health-related outcomes. The identification of states based on morphological characteristics will therefore provide invaluable insights into the microglia-specific functional mechanisms driving treatment effects. The application of clustering analyses enables the detection of groupings within samples reflecting differences in morphological features. Here we propose the application of three custom-created modules to be used within the open-source software CellProfiler. These modules enable the automated detection of clusters present within the sample of microglia, as well as the assessment of the abundance of these clusters across conditions. The application of the analysis is conducted in a highly user-friendly manner, with a user interface integrated into the pipeline, enabling the performance of the analysis with only minimal user input. The workflow thereby includes the conduction of an outlier assessment, followed by hierarchical clustering and k-means clustering and the generation of interactive graphs to determine the number of microglia states present in the sample. Bar plots displaying the abundance of the microglia states across conditions included in the sample will be created. This approach will facilitate faster and more comparable detection of microglial morphological clusters across studies.