engleski

From mathematical chemistry to drug design

The complexity of biological systems is overpowering. To attempt to follow, in details, the interaction between as drug (a microsystem) and an organism (a macrosystem) is very difficult, if not an unachievable task. Modeling of some kind is therefore essential in drug research. There are many different approaches to this problem. We describe one approach to model the bioactivity of a drug. This is the QSAR modeling. QSAR stands for quantitative structure-activity relationship. The QSAR models are based on the empirical observation that these exists a connection between the structure of a molecule and its biological response. We outline a reasonable approach to building up the QSAR model. QSAR models are usually constructed for sets of structurally related molecules whose bioactivities are ordinarily recorded as single numbers. This requires that the structure of a molecule is must also be numerically encoded in some way. Two sets of numbers are then for example statistically analyzed by way of a suitable algebraic expression. The final QSAR model is thus a regression model and one has to be careful about its statistical stability.

drug design ; molecular descriptors ; modeling ; QSAR models

nije evidentirano