Unsupervised machine-learning identifies latent pyrenoid states linked to mitotic remodeling defects and CO2-dependent growth
This study employs an unsupervised machine-learning pipeline to screen *Chlamydomonas reinhardtii* mutants, identifying 17 novel genes essential for pyrenoid integrity and revealing that subtle defects in this biomolecular condensate during mitosis are linked to impaired CO2-dependent growth, including a specific association with the STT7 kinase.