Cellquant: a vibecoder’s guide to image analysis

Quantitative fluorescence microscopy is central to modern cell biology, yet experimentalists lacking computational expertise have limited ability to extract reproducible measurements from the images they collect. Here we present <monospace>cellquant</monospace> , a pipeline to analyze multi-channel fluorescence images that performs cell segmentation, subcellular localization and colocalization analyses, morphometric quantification, and spatial proximity calculations in two or three dimensions. By replacing the requirement for computational expertise with iterative assistance from large language models (i.e., “vibecoding”), <monospace>cellquant</monospace> enables experimentalists with any level of coding experience to quantify imaging data. We validate <monospace>cellquant</monospace> on stress granule formation in human HCT116 cells and nucleolar reorganization across a temperature gradient in budding yeast. Dimensionality reduction of a 45-feature output from 8,309 yeast cells resolved cell state transitions and revealed subpopulations of non-responders. Because the <monospace>cellquant</monospace> interface is text based, the exact command used in any analysis can be recorded and re-executed, generating reproducible quantification with visual quality control and statistical rigor. Code, documentation, and datasets are freely available.