Patterning is a fundamental feature of living organisms, playing a key role in species identification, function, and camouflage. In fruit flies, sensory bristles located on the dorsal thorax are connected to the central nervous system, allowing them to sense their environment. These patterned bristles develop from sensory organ precursor (SOP) cells, which arise through a Notch-Delta signaling mechanism involving direct contact between neighboring or distant cells during the transition from the pupal to adult stage. Current experimental and theoretical studies suggest that signaling filopodia play a key role in regulating long-range Notch signaling. However, existing mathematical models use semiheuristic dynamics of these filopodia and are not accessible to scientists without a strong mathematical and computational background. Moreover, we observe that with respect to the patterns formed by SOP cells, little work has been done to provide a quantitative method to easily detect wild type from perturbed patterns. Here, we use a data-driven model of filopodia dynamics to simulate notch-delta signaling within a user-friendly application accessible to scientists without a strong computational background. Additionally, our application utilizes a machine learning model to classify simulated SOP patterns as either wild type or perturbed based on their clustering and row-organization. Our results show that filopodia that take longer to reach their maximum height tend to produce sparser patterns. This new simulation framework makes it possible for biologists to test and refine hypotheses about bristle patterning in silico.
Olugbenle et al. (Sun,) studied this question.