Metabolism is one of the few key cellular process that is currently amenable to computational modeling on a genome scale, moving towards the holy grail of building an in silico cell. My talk will first introduce genome scale metabolic modeling (GSMM), describing what it is about, what it's good for and its current limitations. I will then describe recent work from my lab developing new computational approaches to harness GSMMs to find new drug targets in aging and cancer. It is based on developing and studying `personalized' models of individual cancer cell lines and clinical samples, and on reversing the Warburg effect to inhibit cancer migration and invasion. Our predictions have been experimentally studied and validated by our collaborators, including the Cohen, Gottlieb, Frezza and Van de Water labs. As time permits, I will discuss the challenges and prospects lying ahead in utilizing GSMMs to study human metabolism.