Malignant gliomas are incurable brain neoplasms with dismal prognoses and near-universal fatality, with minimal therapeutic progress despite billions of dollars invested in research and clinical trials over the last two decades. Many glioma studies have utilized disparate histologic and genomic platforms to characterize the stunning genomic, transcriptomic, and immunologic heterogeneity found in gliomas. Single-cell and spatial omics technologies enable unprecedented characterization of heterogeneity in solid malignancies and provide a granular annotation of transcriptional, epigenetic, and microenvironmental states with limited resected tissue. Previous genomic studies have focused primarily on adult IDH-wildtype glioblastoma. However, many pediatric gliomas and a subset of adult gliomas are defined by mutations leading to global epigenetic dysregulation, such as tumors ostensibly driven by mutations in IDH or core histone proteins. Pediatric and epigenetically-driven gliomas harbor remarkable transcriptional programs, immunologically-distinct microenvironments, and incompletely understood topography (unique cellular neighborhoods and cell-cell interactions). Thus, these tumors are the ideal substrate for single-cell multiomic technologies to disentangle the complex intra-tumoral features, including differentiation trajectories, tumor-immune cell interactions, and chromatin dysregulation. In this dissertation, I investigate the molecular basis of treatment resistance and immune response in these glioma subtypes through computational and experimental approaches. First, I discuss our work investigating the heterogeneity of the pediatric high-grade glioma (pHGG) microenvironment within a newly created multiomic atlas of longitudinal pHGG under standard therapy. This work established a set of pan-pHGG neoplastic cell states and characterized a complex network of tumor-myeloid interactions. I describe how this atlas provides a scalable reference for the broader study of pediatric brain tumors and can be used to nominate potential targeted therapies. Finally, I discuss our work in comparing the global immune response of IDH-mutant and IDH-wildtype adult glioma through single-cell profiling of matched intratumoral and peripheral immune cells. These data provide critical insights into the unique characteristics of pediatric and epigenetically-driven gliomas, highlighting distinct tumor-intrinsic and systemic immune features that may inform the development of genotype-specific chemotherapeutic and immunotherapeutic strategies.