Welcome to CellTune

CellTune is an integrative software to expedite accurate cell classification in spatial proteomics.

Designed for researchers analyzing large cohorts, CellTune streamlines cell classification through a comprehensive and highly optimized human-in-the-loop active learning workflow. By strategically selecting cells with high uncertainty, the platform maximizes the impact of user input and enables the discovery of fine-grained and novel cell types. CellTune advances core capabilities across visualization, gating, annotation, and spatial feature extraction — all within a unified, intuitive, and code-free interface — to efficiently achieve state-of-the-art classification accuracy and resolution, at scale.

To benchmark CellTune, we created CellTuneDepot—the largest fully manually labeled multiplexed imaging dataset to date—comprising over 40,000 cells comprehensively annotated by experts across 30 cell types. In addition, CellTuneDepot contains over 3 million high-quality semi-supervised labels generated with CellTune across a variety of tissues and disease states with accuracy comparable to expert annotation.

Together, CellTune and CellTuneDepot provide a powerful toolkit for classifying cells and accelerating biological discovery in spatial proteomics.

Getting Started

Download & Install
Documentation
Tutorials
CellTuneDepot
Manuscript