Code for interacting with the automated neuron and synapse reconstructions of FANC, the Female Adult Nerve Cord GridTape-TEM dataset. See here for more information about the dataset.
This repo can't be pip installed yet (we'll make this possible at some point) so you may want to add it to your PYTHONPATH so you can import its modules regardless of what folder you're in, e.g. if you cloned this repo to ~/repos/FANC_auto_recon, add export PYTHONPATH=~/repos/:$PYTHONPATH to your shell config file (~/.bashrc or ~/.bash_profile or ~/.zshrc)
Folder descriptions:
annotations: Code for interacting with the annotation framework.example_notebooks: Examples of how to use the code in this repo.FANC_Connectomics_General_Intro.ipynbis a good place to start.images: Code and info related to the EM image data.lm_em_comparisons: Code for comparing EM reconstructions with light microscopy data, e.g. color-depth MIP mask searching.nuclei_prediction: For code related to detecting neuron nuclei/somas.segmentation: Code for interacting with chunkedgraph to access the neuron segmentation data.skeletonization: Code related to skeletonization.synapses: Code related to connectivity.transforms: Code for transforming points between different alignment versions of FANC or between FANC and the Janelia VNC template.volume_meshes: Neuropil compartment mesh files (work in progress).
Generate an API key via https://api.zetta.ai/auth/google/login, then create a text file that contains:
{
"token": "YOUR API KEY"
}
and save that file to ~/.cloudvolume/secrets/cave-secret.json