Unique Asymmetric Branching of Drosophila Neurons Optimizes Temporal Dendritic Computation
Neurons execute a versatile array of computations through a complex interplay of factors, including their morphology and synaptic architecture. Dendritic branching encodes upstream inputs into diverse spike patterns transmitted via downstream axons. While earlier studies highlighted the distinct morphologies and functions of a few representative neurons, the availability of large-scale electron microscopy and fluorescent imaging now enables comprehensive data analysis and further simulations to explore structure-function relationships more broadly. This study investigates the general morphological characteristics of diverse neuron types in the fly model. By employing the Strahler Order (SO) metric, we identified a specific bias towards asymmetry in neuronal branching and further investigated the effect of the asymmetry on computational capabilities. Specifically, symmetric branching enhances coincidence detection capability, whereas asymmetric branching increases input order-selectivity. While certain neurons exhibit extreme symmetry or asymmetry optimized for specific tasks, most neurons strike a balance between these computational strategies. This balance underscores the intricate relationship between neuronal structure and function. In contrast to the wide range of branching symmetries found in random bifurcation models, neurons across different species exhibit species-specific asymmetry, suggesting shared underlying branching mechanisms. Our findings provide a fresh perspective on the exploration of neuronal morphologies and their computational roles.