An Researcher Gathers The Vital Data Required For The Qsar Process

883 WordsMay 11, 20164 Pages
The three steps route is mostly adopted, in which, the first step the researcher gathers the vital data required for the QSAR process. This can be achieved through collecting, or designing a training set of chemicals that can help obtain, taking relevant and quality data. The second step is to choose descriptors that can accurately relate chemical structure to biological activity since not all molecular descriptors are useful for a particular model. The last step is to apply statistical methods that correlate changes in structure with changes in biological activity (Perkins, 2003). Different modelling techniques are adopted to come up with models that vividly describe the relationship between the structure and the property the researcher is interested in. An example of this are but not limited to, multiple linear regression, logistic regression, and machine learning methods. The validation process ensures that the model obtained is appropriate and useful. In fact, only one-fifth of the molecular descriptors is used to avoid over-fitting the model. In most cases, multiple linear regression is adapted to generate QSAR Models. The model is thus expressed as follows: Y=a_0+a_1 ×X_1+a_2 ×X_2+⋯〖+ a〗_n ×X_n Where Y is the response or dependent variable, X_1, X_1, …, X_n are the descriptors used to build the model, and a_1, a_2,…, a_n are the corresponding regression coefficients, a_0 is the constant term of the model. Nevertheless, there is a need to establish whether the model
Open Document