Recently, a new generation of mathematical and computational models of cancer has emerged that have been built in close collaboration with experimentalists and are primarily aimed at understanding aspects of cancer progression and treatment. A key to making these models clinically relevant is that they must utilize patient specific data. The quality and quantity of this data varies wildly across tumor type and covers a wide range of spatial scales (e.g. molecular, histology, imaging, tissue, organ). To further complicate matters most cancers have multiple therapeutic options, some have even 20 or more. This produces a huge therapeutic parameter space with vast numbers of potential drug, dose, and schedule combinations. Therefore a key challenge in our developing field is how to best utilize limited patient data to produce clinically useful predictions that can aid both our understanding of this devastating disease and crucially allow us to treat it better. The purpose of this mini symposium will be to present some of the current mathematical models of cancer that make clinically relevant predictions and specifically utilize patient data. Covering a range of different cancers, spatial/temporal scales and mathematical modeling approaches this minisymposium will illustrate our recent experience integrating clinical and experimental data into mathematical models.