BACKGROUND: There is substantial genetic heterogeneity among glioblastoma tumors from different patients. To develop an accurate depiction of the molecular determinants of response to therapy, we used a large collection of glioma stem cell lines (GSCs) to study the impact of this heterogeneity on response to tyrosine kinase inhibition. Twelve glioma stem cell (GSC) lines, which are representative of TCGA molecular subtypes, were utilized in a high-throughput compound-screening assay to identify drug sensitivities to over 350 tyrosine kinase inhibitors (GSK PKIS1) with varying cellular targets and mechanisms of kinase inhibition. METHODS: Cells were screened in 384-well plates at a density of 1000 cells/well with drug concentrations of either 5 or 0.5 µM. Cell viability measurements were taken using Cell-Titer Glo five days after drug treatment and compared to control wells. Compounds were defined as highly active (>90 % inhibition at 5 µM; >50% inhibition at 0.5 µM), moderately active (>70% inhibition at 5 µM; >30% inhibition at 0.5 µM), or weakly active (>50% inhibition at 5 µM; >15% inhibition at 0.5 µM.) Compounds used in the screen were analyzed for inhibition efficacy against 220 kinases (NanoSyn). RESULTS: Structure-activity relationship analysis identified three compounds that were highly active against eleven cell lines, three more compounds that were highly effective against at least seven cell lines, and seven additional compounds that were highly effective against at least four cell lines. Kinases identified as potential high value targets in selective cell lines include AKT1, PLK1, and IGF1R. Limited responses were seen to EGFR inhibitors. A separate set of compounds targeting TGF-beta, STAT3, PI3K and Notch were used for comparison of drug efficacy. Differential kinase activity amongst the effective compounds has been utilized to identify novel kinases for further study. Ongoing experiments are being performed to correlate the molecular phenotypes of the GSC's by analyzing RNAseq, whole exome, proteomic, DNA methylation, and copy number variation data with drug sensitivity. CONCLUSIONS: Integrating drug sensitivity and molecular phenotype will allow us to provide a complete picture of the underlying molecular determinants of GSC sensitivity and resistance to a panel of targeted kinases with the objective of ultimately designing personalized clinical trials for patients with glioblastoma. SECONDARY CATEGORY: n/a.