A multinational study led by the Garvan Institute unveiled AAnet, an advanced artificial neural network designed to characterize the complex cellular subpopulations within tumors, including triple-negative breast cancer. By analyzing single-cell gene expression data, AAnet captures continuous phenotypic variations, improving on traditional clustering methods. This tool enhances understanding of intratumoral heterogeneity, offering potential to tailor more effective, cell-type-specific cancer therapies and address tumor resistance mechanisms.