Our target diseases with high unmet needs are getting more and more complicated and heterogeneous. Accordingly, to predict clinical efficacy of drugs in the preclinical stages, we should avoid relying too much on phenotypically resemble animal models, but seek more mechanistic, human-predictive models. Thanks to accumulation of publicly-available biomedical information, understanding of pathophysiological mechanisms for both patients and disease animal models can be improved. Systems Biology / Pharmacology are practical Translational Science tools to handle such large-scale biomedical data. For instance, Systems Biology analysis with patient transcriptome data can decipher disease mechanisms, and support creating competitive drug concepts. Quantitative systems pharmacology (QSP) modeling & simulation can provide us with rational clinical efficacy prediction based on the currently-best knowledge. Through the model-based approach, we can continuously and rationally decrease the uncertainties, and identify key research questions to be addressed by both preclinical and clinical researches. I will show some examples for applications of translational science tools to address some research questions in the drug development researches.