Developing predictive molecular maps of human disease through community-based modeling

Abstract

1. The failure rate for drugs in clinical development is still startlingly high despite unprecedented investment in RD many compounds are shown to be safe and to engage the intended target but do not improve the primary indication. This failure stems from the simplistic ways in which we have historically studied potential drug targets for complex diseases and indicates a need for more innovative approaches to identify causal relationships between molecular entities and disease. Biology is rapidly changing and becoming a technology and data-intensive science with the development of new instrumentation to measure various molecular states in greater detail. Herein lays an opportunity to transform our understanding of the molecular underpinnings of disease and develop modeling frameworks that can describe complex systems and predict their behavior. Without these models acting as maps, biologists risk drowning in an ever-growing sea of data. This vision for biology, to use large-scale data to model disease, reflects parallel developments in other scientific disciplines: for example, modeling future trends in climate based on complex meteorological information in atmospheric science. The term fourth paradigm has been coined for this “data intensive” science discovery to distinguish it from empiric, theoretical and computational approaches 3