Precision medicine requires us to utilize the right biomarker characteristics to tailor patients for the right therapies. However, it is very challenging to discover markers for patients tailoring in early phase clinical studies due to the limitation in biological understanding and small sample size. A possible solution is to develop a method that helps connect biology to the clinical benefit. This study proposes a novel approach that jointly accounts for multiple efficacy variables, such as progression-free survival time and tumor size. The technique of Structural Equation Modeling (SEM) or the other general models will be explored and implemented to evaluate multivariate association with candidate biomarkers in totality. An interactive tool programmed in BEACH (an R/Shiny app) will be provided for users to run the proposed methods.