The transformation from normal cells to cancer cells and the maintenance of the malignant state and phenotypes are associated with genetic and epigenetic deregulations, altered cellular signaling responses and aberrant interactions with the microenvironment. This intrinsic complexity of cancer biology calls for novel systems level approaches for understanding tumorigenesis and for developing effective therapeutic strategies. Computational approaches in cancer systems biology embraces this complexity and studies cancer from a systems point of view. In this thesis, we combine various cancer systems biology methodologies including mathematical modeling, statistics and machine learning approaches to study B cell lymphoma biology from two different angles and bring about guidance to treatment strategies. In the first chapter, we present a novel kinetic modeling based computational framework for optimizing combinatorial therapies against chronic active B cell receptor(BCR) signaling in B cell lymphoma. In the second chapter, we describe a Bayesian classifier that’s able to stratify cell of origin subtypes of diffuse large B cell lymphoma based on RNA-Seq data, which is predictive of clinical outcomes. We discuss how above computational analysis provides novel methodology for advising therapeutic strategies.