Reengineering the pharmaceutical industry by crash-testing molecules.

摘要: The recent decline in drug approvals and the increase late-stage failures indicate that ability to generate screen large numbers of molecules has not improved pipeline. Perhaps pharmaceutical industry should follow example automotive agree upon a shared modeling language with vendors academics enable integration predictive computational tools across industry. This will then virtual 'crash-testing' drugs before synthesis, biological testing and, most importantly, clinical trials. represents an ambitiously progressive approach using models for simulating every stage discovery development process. Combining relevant algorithms into grand unified model would prioritization best ideas pursuing program, selecting target or synthesizing molecule. successful application these crash-testing principles by any its current proponents could revitalize so failure is avoided.