Challenges and opportunities for oncology biomarker discovery

► Prospective identification of stratification biomarkers needs efficient interpretation of OMICs data. ► Annotated knowledgebases and text-mining facilitate analysis of OMICs data. ► Standardization of information retrieval and annotation is needed for better biomarker prediction. ► A combined data- and knowledge-driven biomarker prediction approach is recommended. Recent success of companion diagnostics along with the increasing regulatory pressure for better identification of the target population has created an unprecedented incentive for drug discovery companies to invest in novel strategies for biomarker discovery. In parallel with the rapid advancement and clinical adoption of high-throughput technologies, a number of knowledge management and systems biology approaches have been developed to analyze an ever increasing collection of OMICs data. This review discusses current biomarker discovery technologies highlighting challenges and opportunities of knowledge capturing and presenting a perspective of the future integrative modeling approaches as an emerging trend in biomarker prediction. Crisis of pharmaceutical industry prompts Research and Development (R&D) focus from blockbusters to niche busters; its clinical success depends on successful prediction of stratification biomarkers based on combining data and knowledge as an integrative model.