Programmatic use

CAVE is distributed across several cloud-based services that communicate via REST endpoints. There are a number of tools, mostly based around the python ecosystem, that faciliate interaction with a CAVE dataset.

General Tools

CAVEclient

The most important tool for working with CAVE is the CAVEclient, which lets you interact with virtually all aspects of a CAVE dataset. Using CAVEclient, you can find information about datasets, query annotations like synapses or cell types, map the edit history of cells, upload new data, and more.

CloudVolume

CloudVolume is a python client for accessing volumetric data such as imagery or segmentations as well as meshes.

NGLui

NGLui helps build data-driven Neuroglancer states to explore analysis in the context of data.

PCGskel

PCGskel offers one approach to quickly generate skeleton representations of neurons using the PyChunkedGraph topology. In additon, you can add synapses, surface area, radius, and volume information, as well as other labels and annotations to get a rich description of neuronal morphology.

Dataset-specific tools

Several CAVE datasets have provided clients have released clients specifically to work with their data.

FAFBseg

FAFBseg offers both a python client and an R client designed to work with the adult fly brain, including the CAVE-backed Flywire data.

FANC-fly

A python package for working with the ventral nerve cord dataset FANC.

BANC

A python package for working with the whole Drosophila nervous system.

Tutorials

In addition to the documentation for individual tools, several tutorials describe how to work programmatically with CAVE data.

MICrONs Tutorial

MICrONs Tutorials offer a number of instructions for accessing annotations, meshes, skeletons, and more in the mouse visual cortex dataset.

Microns Binder

A collection of notebooks built to introduce working with the MICrONs cubic millimeter dataset.