I built a sandbox for exploring the Drosophila brain connectome using FlyWire data, Python, and Google Colab.
It grew out of a hands-on connectomics workshop I taught at the Technical University of Munich (TUM). The goal was to let people with different neuroscience and programming backgrounds work directly with real connectome data without having to set up the full environment themselves.
The package includes an introductory presentation, a guided Colab notebook, neuronal and connectivity datasets, helper code, and example projects. It covers topics such as cell classes, morphology, connectivity, network structure and synaptic polarity.
During the workshop, students used the same starting material to build their own small projects. AI assistance was available for explaining and modifying code, debugging, and bridging technical gaps, while the questions and analyses were determined by the students.
Beginners can follow the guided workflow, while more experienced users can use the same data and code as a starting point for their own analyses.
Feedback is welcome, and I’d especially be interested to see what questions or projects other people build with it.
AmitGross•46m ago
It grew out of a hands-on connectomics workshop I taught at the Technical University of Munich (TUM). The goal was to let people with different neuroscience and programming backgrounds work directly with real connectome data without having to set up the full environment themselves.
The package includes an introductory presentation, a guided Colab notebook, neuronal and connectivity datasets, helper code, and example projects. It covers topics such as cell classes, morphology, connectivity, network structure and synaptic polarity.
During the workshop, students used the same starting material to build their own small projects. AI assistance was available for explaining and modifying code, debugging, and bridging technical gaps, while the questions and analyses were determined by the students.
Beginners can follow the guided workflow, while more experienced users can use the same data and code as a starting point for their own analyses.
Feedback is welcome, and I’d especially be interested to see what questions or projects other people build with it.