When talking about cancer, a lot of hallmarks can picture it, such as self-sufficiency in growth signals, evading apoptosis, invasion, and metastasis, and so on. However, the driving factor that really defines a cancer is oncogenic mutation, the intrinsic feature of a cancer cell. Over the years, efforts and progresses have been made in targeting either oncogenes themselves or their downstream effectors, but cancer still remains the second leading cause of deaths out of all diseases. More and more studies seek therapeutic opportunities by looking into other cancer vulnerabilities, and the cancer extrinsic factors have gained much attention in the recent years, such as the immune system, tumor microenvironment, and cell-cell interaction. Here, I will present my doctoral research, a study from cancer cell intrinsic factors to extrinsic factors and both directions decipher the molecular mechanisms of a tumor. For cancer intrinsic factor study, we looked at cancer cell metabolic alterations that are oncogene dependent. Here we report that, from an RNAi screen, the top candidate AHCYL1 is overexpressed in human melanoma harboring mutant NRAS but not BRAF or WT. In addition, AHCYL1 is selectively critical for both NRAS-mutated human melanoma cell proliferation and tumor growth. Specifically, we identify AHCYL1 as an oncogene-dependent key regulator of ER calcium homeostasis, with its deficiency leading to decreased ER calcium levels, activating the UPR and ultimately causing cell apoptosis. Our findings suggest that targeting the AHCYL1-IP3R axis presents a novel therapeutic approach for NRAS-mutated melanomas, with potential applicability to all cancers harboring RAS mutations, such as KRAS-mutated human colorectal cancers. For the cancer extrinsic factor study, from a high-content screen using a human blood nutrient library in combination with a fibroblast reporter (FIRE) system, we identified and validated that extracellular ATP (eATP) blunts TGFβ-induced SMA expression and promotes cytokines-induced IL-6 expression in pancreatic stellate cells (PSCs) in both time- and dose-dependent manner. Pretreatment of PSCs with physiological concentrations of eATP is sufficient to alter fibroblast states and to promote PDAC tumor growth. Mechanistically, our data suggest that eATP acts independently of its canonical purinergic receptor pathways to promote an iCAF state in PSCs. The JAK-STAT pathway, one of the major regulators of iCAFs, is strongly activated upon eATP treatment and mediates its effect in inducing an iCAF state. Further, we showed that this is serum dependent. These findings support the idea that TME metabolites can alter PDAC progression by regulating CAF heterogeneity.