A deep learning framework for automated and generalized synaptic event analysis

Mitochondrial dysfunction drives a neuronal exhaustion phenotype in methylmalonic aciduria

Methylmalonic aciduria (MMA) is an inborn error of metabolism resulting in loss of function of the enzyme methylmalonyl-CoA mutase (MMUT). Despite acute and persistent neurological symptoms, the pathogenesis of MMA in the central nervous system is poorly understood, which has contributed to a dearth of effective brain specific treatments. Here we utilised patient-derived induced pluripotent stem cells and in vitro differentiation to generate a human neuronal model of MMA. We reveal strong evidence of mitochondrial dysfunction caused by deficiency of MMUT in patient neurons. By employing patch-clamp electrophysiology, targeted metabolomics, and bulk transcriptomics, we expose an altered state of excitability, which is exacerbated by application of dimethyl-2-oxoglutarate, and we suggest may be connected to metabolic rewiring. Our work provides first evidence of mitochondrial driven neuronal dysfunction in MMA, which through our comprehensive characterisation of this paradigmatic model, enables first steps to identifying effective therapies.

A deep learning framework for automated and generalized synaptic event analysis

Quantitative information about synaptic transmission is key to our understanding of neural function. Spontaneously occurring synaptic events carry fundamental information about synaptic function and plasticity. However, their stochastic nature and low signal-to-noise ratio present major challenges for the reliable and consistent analysis. Here, we introduce miniML, a supervised deep learning-based method for accurate classification and automated detection of spontaneous synaptic events. Comparative analysis using simulated ground-truth data shows that miniML outperforms existing event analysis methods in terms of both precision and recall. miniML enables precise detection and quantification of synaptic events in electrophysiological recordings. We demonstrate that the deep learning approach generalizes easily to diverse synaptic preparations, different electrophysiological and optical recording techniques, and across animal species. miniML provides not only a comprehensive and robust framework for automated, reliable, and standardized analysis of synaptic events, but also opens new avenues for high-throughput investigations of neural function and dysfunction.