Presentation description

Identifying multiple myeloma patients at high risk of early treatment failure is a critical need. Most patients fail first-line therapy and ~50% of patients do not survive past 5 years after diagnosis. This project shows that SPECTRA, a statistical technique developed by the Camp Lab, can characterize global gene expression by representing it as multiple quantitative tumor variables. Spectra variables allow gene expression to be incorporated into predictive modeling to identify high-risk groups.