Corvo is a tool for the visualization of dimensionally reduced single-cell transcriptomics datasets in virtual reality. Built with the scenegraphing and rendering library Scenery (https://github.com/scenerygraphics/scenery), both Corvo and its underlying infrastructure are open-source.
Corvo is designed to be launched from CorvoLauncher (https://github.com/ljjh20/CorvoLauncher) with the accompanying data pre-processor and UI.
For development, clone the repository and launch with a JVM-compatible IDE.
git clone https://github.com/ljjh20/Corvo.git
Corvo targets JDK-11. Please set the SDK and language level (Project Structure in IntelliJ) to version 11 (e.g. from AdoptOpenJDK or Temurin Eclipse).
You may want to dedicate additional RAM, using e.g. -Xmx16g for 16GB, in the VM options.
A large repository of compatible single-cell transcriptomics datasets are available from the Chan Zuckerberg Initiative: https://cellxgene.cziscience.com/datasets
However, these must be pre-processed with the PreProcess utility in CorvoLauncher:
You will need to download a Vosk English language model from: https://alphacephei.com/vosk/models and place it in the repository root.
Corvo is launched by running main(arrayOf(<dataset path>, <language model path>)) in the companion object of src/main/kotlin/graphics/scenery/corvo/XVisualization.kt.
Feature list to be added soon