Methods of applying QSAR to predict In vivo and In vitro activity Relationship Paradigm
Author(s): Bhuvnesh Rai, Medha Srivastava, Dharmendra Kumar Chaudhary
The Chemical structure is correlated by QSAR with an activity relationship using many statistical approaches. It is the model used for the various purposes as a prediction of activities in in-vivo and in-vitro activities of molecules which are not being tested chemically. These efforts have been simulated by scientists for a there wider range of applications as toxicology, etc. Algorithms available for generating Quantitative Structure-Activity Relationship (QSAR) are single biological activity based on multiple regressions with molecular descriptors of a data set. Such an analysis is providing correlation only with a specific biological activity either in vitro or in vivo or specific target thus limiting their use in only one environmental condition. If one would like to use an integrated approach for comparison of activity measured in vitro with that of in vivo, the current methods have limitations. “The objective of my work is to identify the potential descriptors set explaining activities in vitro and in vivo. PLS regression was employed to predict the in vitro and in vivo activity using the set of potential descriptors. This procedure yielded improved predictability of biological activity from potential molecular descriptors. QSAR models are scientifically given credibility as the tool for prediction and classification of activities of untested molecules biologically and chemically.