Why do you need a computer to design a drug?

Two problems are commonly recognized.

• An increase in the time of entering a new drug to the market. It took about 8 years to bring a New Chemical Entity (NCE) into the market for compounds that entered clinical trials in the 1960s, and this process took about 15 years in 1997.
• An increase in the cost of getting a new drug onto the market. One needed to make an investment of about 150 USD for a drug development in the 1960s, and it required approximately 600 USD for the development process in 1997.

Why do they grow? A rationalization is that investments shifted from acute (thereby, simpler) illnesses to chronic and degenerative illnesses (thereby, more complex). Another explanation is that clinical toxicology and safety requirements continuously toughen.

It seems hardly possible to influence the stage of clinical investigations, but it can be real to affect the discovery research stage, which takes 10 years (in other words, 2/3 of the total drug discovery process) and about 1/2 of the total cost.

How to reduce them? A solution is called the "rational drug design" or "computer-assisted drug design" or "computer-aided drug design". By some estimation, the expected long-term impact of these technologies will be as much as 25 percent as success rates increase and the total time required to bring a product to market falls.

The rational drug design is the design of drugs guided by information. The information includes knowledge about targets (usually proteins) and ligands (usually small molecules).

Currently, ChemoSoft has tools to start with the second information type (structures of ligands). It contains both instruments for hit generation or enrichment of an in-house collection (various utilities for the sorting and selection by diversity [including Diversity of Heterocycles]) and ones for hit-to-lead optimization and lead development (the selection by similarity [including Bioisoster similarity], clustering, computation of ADME-relevant parameters and tools for QSAR studies).

Diversity of Heterocycles, a resulting plot.	Computation of ADME-relevant parameters.
QSAR. Neural network in action. Calculation results.	Similarity of Biosters. An analogue of Bayer® 's aspirin® found.