Advantages and Disadvantages of QSAR

ADVANTAGES OF QSAR

• It gives quantifying the relationship between structure and activity with their physiochemical property basis.

• Possible to make predictions of designed compounds before the chemical synthesis of novel analogues.

• It may help to understand the interactions between functional group of designed molecules and their activity of target enzyme or protein.

DISADVANTAGES OF QSAR

• Due to biological data experimental error it may give false correlations.

• If training set of molecule is less, the data may not reflect the complete property and it cannot be used to predict the most active compounds.

• In some 3D QSAR study ligands binding receptor or protein may not be available in that case the

common approach result may not represent the reality.

• Cannot expect that the QSAR works all the time give successful applications.

A model perfectly predicts that the training data may not be good or even useless for prediction. The problem of QSAR is to find coefficients C0, C1 , ... , Cn such that

Biological activity = C0 + (C1 × P1) + ... + (Cn × Pn) 

and the prediction error is minimized for a list of given compounds.

Partial least squares (PLSs) is a technique used for computation of the coefficients of structural descriptors.