A new open-source R package, CRAWDAD, is transforming the analysis of cell-type spatial relationships within tissues, utilizing the powerful capabilities of spatially resolved omics (SRO) data. This innovative tool is anticipated to provide significant insights about how different cell types organize and function within various tissues, directly impacting our comprehension of both healthy physiology and disease progression.
Spatially resolved omics technologies allow for the molecular profiling and identification of specific cell types, all the meanwhile preserving their physical organization within tissues. Evaluations of cell-type spatial relationships, namely colocalization and separation, offer researchers the opportunity to explore intercellular dynamics, influencing factors related to tissue functionality and pathologies. Leveraging these advanced methodologies, CRAWDAD aims to cater to the growing need for precise quantitative analysis of these spatial dynamics.
Developed by the research team at JEFworks Lab, CRAWDAD (Cell-type Relationship Analysis Workflow Done Across Distances), stands out by enabling evaluations of multi-scale spatial relationships across diverse datasets. The package employs sophisticated statistical techniques, primarily centered around binomial proportion testing and Z-scores, to determine if observed mixtures of cell types within specified neighborhoods differ significantly from random distributions. This comparative framework creates opportunities for assessing relationships not only within localized spatial contexts but also across broader scales.
“The multi-scale analysis of SRO data enabled by CRAWDAD will provide useful quantitative metrics to facilitate the identification, characterization, and comparison of cell-type spatial relationships,” said one author of the research, emphasizing the capabilities of the tool.
CRAWDAD’s utility has been demonstrated through multiple case studies, applying the software to simulated datasets as well as real-world tissue samples from sources including mouse cerebellum, developing embryos, breast cancer, and human spleen. For example, analyses of mouse cerebellum tissue—organized layers distinctively contributing to motor control—revealed expected colocalization trends between Purkinje neurons and Bergmann glia, which are known to physically interact closely. Such results corroborated the accuracy and relevance of CRAWDAD’s evaluations, showcasing its effectiveness against existing spatial analysis methods.
Beyond characterizing these relationships, CRAWDAD reveals how cell-type dynamics may manifest uniquely depending on tissue architecture and composition. This became evident when analyzing corresponding datasets from multiple samples; CRAWDAD identified variations and similarities among cell-type spatial dynamics based on specific conditions, such as health, disease, or developmental stage. “Such multi-scale characterization enabled by tools like CRAWDAD can reveal distinct insights not apparent in evaluations of spatial relationships,” added the authors.
The CRAWDAD package is now available for public use as open-source software, opening new avenues for researchers worldwide to apply its multi-scale analysis capabilities across diverse tissues. Enthusiasts can explore it at https://github.com/JEFworks-Lab/CRAWDAD, where comprehensive documentation and tutorials accompany the software.
The implementation of CRAWDAD may serve as the precursor to comprehensive atlasing efforts, fostering the integration of tissue-based analyses and potentially informing diagnostic and treatment strategies. By employing quantitative spatial trend metrics, CRAWDAD could immensely contribute to research hubs intending to understand the intricacies of cellular organization within the body.
Finally, as researchers continue to refine spatially resolved omics technologies and analyses, CRAWDAD is expected to evolve as well, adapting to new datasets and methodologies to remain at the forefront of spatial analysis. The hope is to pinpoint associations between cell-type distributions and their physiological roles, leading to breakthroughs not only within biomedical research but also across therapeutic landscapes.
“CRAWDAD can quantitatively capture asymmetric cell-type spatial relationships and effectively delineate cell-type spatial relationships across multiple length scales for diverse tissues and SRO technologies,” concluded the authors, encapsulating the broad potential applications and future directions of this promising tool.