CryoViT Documentation

Welcome to the CryoViT documentation!

CryoViT is a deep learning library for automated segmentation of Cryo-ET data. It provides both a command line interface (CLI) and a napari plugin for easy use, as well as tools for training and evaluating custom models.

More details can be found in the preprint, and you can check out the source code for implementation details.

Hint

• If you already have a dataset, and just want to get it segmented using one of the provided pre-trained models, check out the quick start guide.

• If you are new to CryoViT and want to train your own segmentation model, check out the user guide.

• For more information on running experiments like those shown in the CryoViT paper, check out the GitHub repository for example configuration files and instructions.

• If you run into any issues, check out the help section to see if your issue has already been addressed.

Getting Started

These sections cover the basics of getting started with CryoViT, including installation instructions for use and development, setting up dependencies like napari, and a quick start guide to training and using your first model.

Getting Started

• Installing CryoViT
  • PyPI package
  • Installation from Source
• Setting up Napari
  • Installing Napari
  • Installing the CryoViT Napari Plugin
• Quick Start Guide
  • Viewing Tomogram Data
  • Viewing Model Information
  • Running Inference Script
  • Viewing Segmentation Results

User Guide

These sections go over the basic workflow of using CryoViT, from preparing your dataset for training or inference, to running commands using either the CLI or napari plugin.

User Guide

• Setting up a dataset
  • Dataset Structure
  • Label Formats
  • Using Napari
• Running from the command line
  • Training a Model
  • Using a Pre-trained Model
• Running using Napari
  • The Training Manager
  • The Inference Manager

Reference

These sections contain the API reference for CryoViT, for those seeking to expand or customize the library, or simply to dive deeper.

cryovit.run	Functions to run feature extraction, training, evaluation, and inference for users.
cryovit.models	Implementations of deep-learning automated segmentation models for CryoViT.
cryovit.datasets	Implementations of PyTorch datasets for loading Cryo-EM tomograms.
cryovit.datamodules	Implementations of PyTorch Lightning DataModules for loading Cryo-EM tomograms.
cryovit.types	Custom types and dataclasses for CryoViT models.
cryovit.config	Hydra configuration classes for CryoViT experiments.
cryovit.utils	Utility functions to process data and models in a format recognizable by CryoVIT.