Background: Synthesis of organic compounds with specific biological
activity or physicochemical characteristics needs a thorough analysis of the enumerable
data set obtained from literature. Quantitative structure property/activity relationships
have made it simple by predicting the structure of the compound with any
optimized activity. For that there is a paramount data set of molecular descriptors
(MD). This review is a survey on the generation of the molecular descriptors and its
probable applications in QSP/AR.
Methods: Literatures have been collected from a wide class of research journals, citable
web reports, seminar proceedings and books. The MDs were classified according
to their generation. The applications of the MDs on the QSP/AR have also been
reported in this review.
Results: The MDs can be classified into experimental and theoretical types, having a sub classification
of the later into structural and quantum chemical descriptors. The structural parameters are derived from
molecular graphs or topology of the molecules. Even the pixel of the molecular image can be used as
molecular descriptor. In QSPR studies the physicochemical properties include boiling point, heat capacity,
density, refractive index, molar volume, surface tension, heat of formation, octanol-water partition
coefficient, solubility, chromatographic retention indices etc. Among biological activities toxicity, antimalarial
activity, sensory irritant, potencies of local anesthetic, tadpole narcosis, antifungal activity, enzyme
inhibiting activity are some important parameters in the QSAR studies.
Conclusion: The classification of the MDs is mostly generic in nature. The application of the MDs in
QSP/AR also has a generic link. Experimental MDs are more suitable in correlation analysis than the
theoretical ones but are more expensive for generation. In advent of sophisticated computational tools
and experimental design proliferation of MDs is inevitable, but for a highly optimized MD, studies on
generation of MD is an unending process.