|
|
|
|
|
|
|
stations for graphical visualization. At the time of this writing, Merck, Bristol-Myers Squibb, Pfizer, Dupont, Eli Lilly, and Hoechst, all have Cray supercomputers. Many others have substantial multiprocessor work stations. |
|
|
|
|
|
|
|
|
The cost of hardware varies with need. A biotech firm that requires only protein homology building of a limited number of proteins and employs a single protein structure specialist could probably make due with one of the workstations mentioned above and so incur initial hardware costs of 100,000 USD (not including yearly maintenance fees). However a company with many diverse projects and 10 or more computational specialists will require considerably more and can incur computer hardware purchases of 510 million USD and many hundreds of thousands in yearly maintenance agreements. |
|
|
|
|
|
|
|
|
CADD has already made an impact and is integrated into successful research programs in many companies. However, although the successes of CADD have been significant, they at times seem modest. This is understandable since CADD (and most of its constituent fields) is a young pursuit. Much of it is still in development and retains an aura of basic research. Powerful computers, the engines of these methods, have only recently been routinely available. Also, in some respects the development of CADD has paralleled the development of the techniques of macromolecular structure determination (protein NMR and crystallography), which have also only recently become routinely and widely used. Chemical software, the other principal tool of CADD can do no more than track the progress of its underlying science. Also, mostly because of limited computer power, computational solutions always carry with them a certain level of approximation. |
|
|
|
|
|
|
|
|
Luckily, however, diminution of these shortcomings is occurring rapidly and CADD techniques can become only more powerful and accessible. Computer hardware is becoming faster and cheaper. This will alleviate the need for many of the risky approximations that are currently employed. Also, the last few years have seen some impressive strides in the power and ease of use of CADD software. Software developers are continually working to develop better integrated and more user-friendly programs. Perhaps most important, the sciences that underlies CADD are intense areas of research that are constantly advancing and becoming better understood and more refined. Improved force fields will provide more accurate estimates of the interactions between proteins and drug molecules. The often ignored role of solvation and ionic strength in molecular conformation and biological interactions are being incorporated into calculations. New quantum chemical methods are under development that will provide more rapid determination of structure and properties. Understanding of protein structure and function is an intense field with daily advances. |
|
|
|
|
|