Introducing dendrites to spiking neural networks
Although neuronal dendrites play a crucial role in shaping how individual neurons process synaptic information, their contribution to network-level functions has remained largely unexplored. Current spiking neural networks (SNNs) often oversimplify dendritic properties or overlook their essential functions. On the other hand, circuit models with morphologically detailed neuron representations are computationally intensive, making them impractical for simulating large networks.
In an effort to bridge this gap, we present Dendrify—a freely available, open-source Python package that seamlessly integrates with the Brian 2 simulator. Dendrify, through simple commands, automatically generates reduced compartmental neuron models with simplified yet biologically relevant dendritic and synaptic integrative properties. These models offer a well-rounded compromise between flexibility, performance, and biological accuracy, enabling us to investigate the impact of dendrites on network-level functions.
If you use Dendrify for your published research, we kindly ask you to cite our article:
Pagkalos, M., Chavlis, S., & Poirazi, P. (2023). Introducing the Dendrify framework for incorporating dendrites to spiking neural networks. Nature Communications, 14(1), 131. https://doi.org/10.1038/s41467-022-35747-8
Documentation for Dendrify can be found at https://dendrify.readthedocs.io/en/latest/
The project presentation for the INCF/OCNS Software Working Group is available on google drive.