Two distinct approaches are possible in the area of computer-aided drug
design. If the molecular structure of the target macromolecule is known the methods
are obvious and direct and have achieved a high level of sophistication. That area
may be extended by using computational techniques to predict protein structure as
illustrated here by the interleukin receptor.
When the only lead is a set of known active compounds or knowledge of a
biochemical transformation which is to be interrupted, then the path is less direct.
Currently favoured tactics include the use of molecular similarity methods and the
employment of neural networks. Recent advances include the prediction of the
relative potency of different chiral forms of drugs.
Link to this comment:
All Comments (1)