ABSTRACT
Accurate models that are able to assess the distinct disease states of patients and predict not an average, but an individual, response to specific drugs are essential keystones of customized therapy.
The convoluted nature of organisms and the resulting complexity of data—in particular the inherent high dimensionality—still pose significant challenges to a broad range of applications in personalized medicine. Even so, the increasing accumulation of data and recent advances in data analysis and modeling technologies for biological systems already offer some promising opportunities.
This section discusses different current approaches ranging from mechanistic modeling concepts to data-driven modeling techniques and analyzes their respective strengths and shortcomings with regard to their applicability in the context of personalized medicine.
Special emphasis is placed on the importance of emerging multi-scale approaches that reflect the hierarchical architecture of organisms and enable the systematic investigation of both short-term therapeutic effects and possible long-term adverse reactions.
Moreover, the potential of integrative approaches that incorporate a priori mechanistic knowledge and data-driven algorithms into a common framework is examined.
