PanoMito: mapping mitochondrial heterogeneity from morphofunctional profiling to single-mitochondrion spatial transcriptomics
The paper introduces PanoMito, an integrated toolbox combining a large multi-modal mitochondrial dataset and a deep-learning framework to enable high-precision segmentation, unsupervised classification, and single-mitochondrion spatial transcriptomics, thereby establishing a comprehensive framework for systematically mapping mitochondrial heterogeneity and morphology-function relationships across diverse species.