Funded project, 2024 Biological computer for neuronal networks in Health and Disease
Neurobiological computing and Synthetic Intelligence
Project Outcome
The goal of this funded project (Explore!Tech Grant by Health + Life Science Alliance Heidelberg Mannheim) was to develop a prototype GUI “neuroseq”, which allows a user to edit and simulate in silico models of an assembly of neurons with different neuronal architectures and model parameters.
The resulting software used PyQT for the GUI elements, PyQtGraph for the interactive plotting components, and the Brian 2 simulator for simulating the spiking neural networks. The stimuli can be edited with a user-friendly interface that is similar to a MIDI music editor, and as well as trace and spike graphs of the simulation results, a heatmap of the neuronal connectivity as well as custom analysis plots of selected regions can be displayed. There is also a built-in Python console to allow the user to interact programatically with their model, for example to use numpy to create or modify the arrays of stimuli.
It was a great experience working as a team with Liam and Harald on the software development project. Their expertise, collaboration, and dedication made the process smooth and enjoyable. I truly appreciated the synergy we had throughout the project and am looking forward to future opportunities to work together again.
Dr. Andrey Formozov, Neurophysiology, Medical Faculty Mannheim
Enabling new research
Through this prototype, the researcher was able to plan and apply for further research grants to support his research. With the prototype, it also became clearer as to which approach should be pursued in preparing and saving model representation of neuronal architectures, and what kind of user interface was appropriate for the different tasks. Further software development is planned, such as integrating the software with a holographic stimulation system to control the optogenetic excitation of neurons, and compare these real world results to the in silico simulation results, making this prototype a foundation for future research directions.
Resulting publications
- Bernstein Conference 2024, Frankfurt-am-Main, 09/2024
NeuroSeq GUI
