by Richard G. Lanzara

Created on: May 08, 2008

Currently most drug research and development requires computing of some sort from the modeling of the biological response to the statistical effects of drugs on animals and humans. This is a very large and complex area, which can only be briefly introduced in this essay. Computing requires the mathematical modeling of several critical areas including an understanding of how drugs work at their target receptors to determining the proper dosages that should be administered to people. Although many scientists know only their relatively narrow view of the overall computational landscape in pharmaceutical drug research, these computational areas are vitally important if we are to have safer and more effective drugs in the future.

Finding suitable molecules that can react with the target receptors on our cells in a specific way, requires a range of scientific expertise that includes molecular modelers, synthetic chemists, systems biologists, toxicologists, statisticians and mathematicians. Each scientist uses specific computations within their areas of expertise to solve these problems in drug research and development. Usually this work also includes computational models that predict how biological and chemical systems function when exposed to new molecules that may ultimately become FDA approved drugs. Even today this is a fundamental problem in drug research and development that is reflected in a profusion of companies that assist the major pharmaceutical companies with their own computational software to model how potential drug molecules affect biological systems. How drug molecules target and activate cellular receptors is a very active area of research that remains a fundamental and unresolved problem today.

Although computing implies a direct knowledge of biological systems, a complete mathematical characterization is extremely complex and will require much more scientific development. Often, it is difficult to determine the number and type of unknown variables to include within a computational model for any particular biological system. These unknown variables should relate to known physical entities in order to have a logical consistency with known biophysical laws, but this is sometimes ignored if the model mostly fits the available data. Ignoring this aspect, however means that the computational model may no longer have logical underpinnings to the physical world.

One way computational models are systematically developed and tested is by determining their

• 1
• 2
• Next Page >>