Zhao et al. utilize a multi-optic system to investigate the role of lactate metabolism in cutaneous melanoma (SKCM). Additionally, they propose a model to predict patient outcomes and identify RAB32 as a potential therapeutic target. The researchers utilize CellProfiler and a pre-trained AI model to extract key features from multiple types of biological data, including gene expression, tumor imaging, single-cell sequencing, and spatial transcriptomics. They measured lactate metabolism scores and examined whether specific cell types within the tumor were responsible for driving lactate metabolism. They tested 101 different model combinations from 10 algorithms to build the best possible predictor model. The study found that melanoma cells exhibit the highest lactate metabolic activity, as determined by analyzing 443 images. A Lactate Metabolism Signature (LMS) model was then created to predict patient outcomes. Patients in the high LMS group had a shorter life span, suspected immune evasion, and activated tumor pathways. On the other hand, patients in the low LMS group had better survival rates, more immune infiltration, and a higher response to immunotherapy. The results also showed that RAB32 is a critical gene, as it is highly expressed in tumors and is associated with poor patient outcomes.