Heterogeneity of tumors including the presence of resistant clones represents a great challenge for high risk neuroblastoma (NB) and malignant pheochromocytoma (PCC). Currently, there are no treatments that address cancer heterogeneity. Classification of bulk tumors and corresponding predictions are poor and too general. Understanding heterogeneity will provide critical insight for the development of targeted therapeutic strategies. Here we elucidate the origin and heterogeneity of neuroblastoma and PCC. We will develop a new analytic and experimental approach based on single nuclei transcriptomics for predicting promising treatment strategies that take into account tumor heterogeneity. Aim: Using single nuclei transcriptomics of human NB and PCC tissue to explore heterogeneity, cell of origin and immaturity. Identifying the pathways that can drive tumors into differentiated state might unlock new promising cancer treatment strategies. Significance: Knowing the pathways and target genes that can drive tumors into less metastatic and differentiated states can be a key for cancer treatment. To achieve this, we need to understand heterogeneity and “differentiation-like” status of heterogeneous cancer cells as compared to their normal origin cell types. We also need to know the differentiation path of normal cancer-originating cells in terms of expressed genes. Here, we will use human neuroblastoma tumors and mouse model systems to develop such comparative differentiation trajectory analysis, to make predictions and, finally, to perform preliminary validations of envisioned tumor-differentiation strategies.