Whole genome doubling (WGD) facilitates rapid tumor evolution and is a hallmark of many cancers, including Glioblastoma (GBM) where it has an incidence of ~14%. GBM remains uniformly lethal, with a median overall survival of 14-16 months, despite aggressive therapy. Prior studies indicate that the outcome of the competition between WGD+ cells and their diploid ancestor cells sets a crucial branching point in the evolution to malignancy, and that resource access is in prime position to determine that outcome. However, the exact nature of these interactions between cell genotype and the tumor microenvironment (TME), how these interactions can predict the GBM aggressiveness and therapy response, and potential reversal of GBM growth by genotype specific TME manipulation are key questions that remain to be investigated. Here, we present a calibrated mathematical model that can predict the growth and invasion of a glioma, given its (poly)aneuploidy and the nature of its TME. Several computational models that simulate GBM growth as a function of the brain microenvironment, using reaction diffusion PDEs, have been previously developed. We build upon this body of work to develop a new model whose assumptions in part overlap with those made by the previously developed PIHNA model for glioma-microenvironment interactions. The model presented here addresses three challenges prior mathematical models of GBM invasion have been grappling with – the integration of multiple data types, the impact of intra-tumor heterogeneity and model evaluation in human patients. The model fundamentally tracks the evolution and interaction of various cell genotypes, nutrients and TME physical properties using PDEs that incorporate generation, removal, and transport. However, with a focus on computational efficiency and cost, the model is solved by discretizing the brain into voxels, and by focusing on solving simplified equations governing the state of each individual voxel. The state of each voxel is defined by 8+ variables that are updated over time: stiffness, oxygen, phosphate, glucose, vasculature, dead cells, migrating cells and proliferating cells of various ploidies, and treatment conditions such as surgery, radiotherapy, and chemotherapy. We dub this model as the Stochastic State Space Model of the Brain (S3MB). S3MB allows for rapid computation of GBM growth within the brain with patient specific clinical and physiological data as input and temporal evolution of GBM cell communities over several months as output (<5s of computation time for >12 months of GBM growth). The model can be calibrated using in vitro, in vivo, and clinical patient data and used to infer inter-patient variability in the conditions that dominated at the onset of a patient’s glioma formation and comparing them to the respective aneuploidy status of the tumor at the time of detection. The model is currently being applied towards classifying TMEs by their propensity to select for or against WGD. Understanding the selective pressures that shape ploidy at the onset of glioma formation will help us mimic their reverse therapeutically after tumor detection.