Quantifying bristle cell organization in

Developing tissues reproduce nearly identical patterns from animal to animal, despite the process being stochastic at the cell level. A major challenge for researchers is the ability to quantify and classify complex cell and tissue spot patterns across wild type and perturbed conditions. Here, we use the organization of small sensory bristles on the fruit fly thorax as a model system to address this problem. A well-known and easily observable distinguishing feature of spot patterns is density. Beyond this, it is unclear how to quantitatively distinguish between patterns in a reproducible way. Our work evaluates the utility of the spatial clustering of spot patterns in quantifying their organization. We propose four clustering features, obtained using the density-based spatial clustering of applications with noise (DBSCAN) algorithm. Together with pattern density we assess how these features can quantify and distinguish between bristle patterns in a variety of wild-type and mutant flies to confirm known perturbations and discover new ones. Our results show that a combination of spot pattern density and the variance between the size of clusters is sufficient to distinguish between 70% of patterns from wild-type and mutant tissues.