Biomedical Engineering Reference
In-Depth Information
Drug Discovery
Pharma, the primary backer of bioinformatics R&D worldwide, is keenly interested in automating and
speeding the drug discovery and development process. The typical drug discovery and development
process, shown in
Figure 9-2
, involves an often arduous series of events that starts with perhaps
5,000 candidate drug molecules and ends with a single product that can be brought to market.
Figure 9-2. The Drug Discovery and Development Process (top) and
Application Areas of Modeling and Simulation (bottom).
Because any technology that can shorten the discovery and development process has the potential to
save the industry billions of dollars, there is considerable R&D involved in replacing or supplementing
the drug discovery process with modeling and simulation. A better understanding of the underlying
metabolism of a particular disease or condition can suggest which molecules will be most effective for
treatment, and which ones may cause toxic reactions in a patient. Similarly, assuming that protein
molecules with similar structure also have similar function, modeling protein structure and comparing
it with known drugs can potentially serve as a more effective screener for candidate drugs, compared
to wet-lab techniques.
Later in the drug discovery process, modeling and simulation of pharmacokinetics and of drug
absorption can potentially be used to shorten clinical trials. Currently, each phase of the clinical trials
takes a year or more. Phase I, involving about 100 subjects, deals with safety. Phase II, which
involves about 200 subjects, deals with evidence for efficacy at various dosages. Phase III, involving
up to about 5,000 subjects, deals with assessing the clinical value of a molecule. Phase IV, which
