Generalizable Machine Learning in Neuroscience using Graph Neural Networks
This work addresses the challenge of achieving generalizable machine learning in neuroscience, particularly for microscopic neural systems, though it appears incremental by building on existing graph neural network methods.
The study tackled the problem of applying deep learning to microscopic neural dynamics and emergent behaviors using calcium imaging data from C. elegans, showing that neural networks perform well on neuron-level dynamics prediction and behavioral state classification, with graph neural networks outperforming structure-agnostic models and excelling in generalization to unseen organisms.
Although a number of studies have explored deep learning in neuroscience, the application of these algorithms to neural systems on a microscopic scale, i.e. parameters relevant to lower scales of organization, remains relatively novel. Motivated by advances in whole-brain imaging, we examined the performance of deep learning models on microscopic neural dynamics and resulting emergent behaviors using calcium imaging data from the nematode C. elegans. We show that neural networks perform remarkably well on both neuron-level dynamics prediction, and behavioral state classification. In addition, we compared the performance of structure agnostic neural networks and graph neural networks to investigate if graph structure can be exploited as a favorable inductive bias. To perform this experiment, we designed a graph neural network which explicitly infers relations between neurons from neural activity and leverages the inferred graph structure during computations. In our experiments, we found that graph neural networks generally outperformed structure agnostic models and excel in generalization on unseen organisms, implying a potential path to generalizable machine learning in neuroscience.