Glioblastoma (GBM), one of the most common and most lethal types of brain cancer, is characterized by a high level of intra-tumoral genetic heterogeneity, which we posit underlies the cancer's extreme aggressiveness and unpredictable therapeutic responses. Treatment development is limited by the inability of current in vitro models to capture this heterogeneity and by the cost, time, reproducibility, and complexity that is inherent to all in vivo models. This proposal outlines a plan to generate a single-cell RNA sequencing database of information on the cellular compositions of primary tumors and their experimental model derivatives. Our lab has developed a 3D tissue-engineered model in which GBM cells are cultured on scaffolds fabricated from hyaluronic acid (HA), an extracellular matrix component that is abundant in normal brain tissues, overexpressed in GBM, and known to promote tumor growth, invasion, and therapeutic resistance. Our overall hypothesis is that HA scaffolds can maintain more cell populations of the primary tumor than the current best in vitro and in vivo models, across all GBM subtypes. Our long-term goal is to tissue-engineer a construct upon which primary tumor cells can be cultured in a manner that preserves their genetic heterogeneity, such that high-throughput screens can be applied to identify patient-specific treatments within a clinically actionable timeframe.