The scientists, and the researchers around the globe generate tremendous amount of information everyday; for instance, so far more than 74 million molecules are registered in Chemical Abstract Services. According to a recent study, at present we have around 1060 molecules, which are classified as new drug-like molecules. The library of such molecules is now considered as ‘dark chemical space’ or ‘dark chemistry.’ Now, in order to explore such hidden molecules scientifically, a good number of live and updated databases (protein, cell, tissues, structure, drugs, etc.) are available today. The synchronization of the three different sciences: ‘genomics’, proteomics and ‘in-silico simulation’ will revolutionize the process of drug discovery. The screening of a sizable number of drugs like molecules is a challenge and it must be treated in an efficient manner. Virtual screening (VS) is an important computational tool in the drug discovery process; however, experimental verification of the drugs also equally important for the drug development process. The quantitative structure-activity relationship (QSAR) analysis is one of the machine learning technique, which is extensively used in VS techniques. QSAR is well-known for its high and fast throughput screening with a satisfactory hit rate. The QSAR model building involves (i) chemo-genomics data collection from a database or literature (ii) Calculation of right descriptors from molecular representation (iii) establishing a relationship (model) between biological activity and the selected descriptors (iv) application of QSAR model to predict the biological property for the molecules. All the hits obtained by the VS technique needs to be experimentally verified. The present mini-review highlights: the web-based machine learning tools, the role of QSAR in VS techniques, successful applications of QSAR based VS leading to the drug discovery and advantages and challenges of QSAR.

Simplified molecular input line entry system (SMILES) notations of 116 π-conjugated organic compounds have been used in three random splits to develop single optimal descriptor based quantitative structure–property relationships (QSPR) models for the prediction of dielectric constants by CORAL (CORrelation And Logic). Four kinds of optimal descriptors were obtained based on SMILES, hydrogen suppressed graph (HSG), graph of atomic orbitals (GAO) and hybrid descriptors. The Monte Carlo optimization was carried out for each random split by three different methods: (i) classic scheme; (ii) balance of correlations; and (iii) balance of correlations with ideal slopes. The QSPR models gave reliable and accurate values of dielectric constant for all the π-conjugated organic compounds. SMILES and the hybrid-based QSPR model provided the best accuracy for the prediction of dielectric constants. Statistical characteristics of the QSPR model-1 based on classic scheme method are n = 110, r² = 0.860, Q² = 0.860, s = 1.84, MAE = 1.30 and F = 696 (training set), n = 6, r² = 0.947, Q² = 0.876, s = 0.955, MAE = 0.647 and F = 71 (test set). These QSPR models are further validated by an external validation set of 25 molecules and the robustness is checked by parameters like k, kk, r_m², r^*_m², average r_m², ∆r_m²and randomization technique ( ).

CORAL software has been used to build quantitative structure–activity relationships (QSARs) for the prediction of binding affinities (pEC₅₀, i.e., minus decimal logarithm of the 50% effective concentration) of 35 potent inhibitors towards the voltage-gated potassium channel subunit Kv7.2. The pEC₅₀ has been modelled using eight random splits, with the following representations of the molecular structure: (i) hydrogen-suppressed graph (HSG); (ii) simplified molecular input line entry system (SMILES); (iii) graph atomic orbitals (GAOs) and (iv) hybrid representation, which is HSG together with SMILES. These models have been examined using three methods, the classic scheme, balance correlation, and balance correlation with ideal slope. The QSAR model based on single optimal descriptors using SMILES provided the best accuracy for the prediction of the pEC₅₀. The robustness of these models has been checked using parameters such as r_m², r^*_m², , and using a randomization technique. The best QSAR model based on single optimal descriptors has been applied to study the in vitro structure–activity relationships of pyrazolo[1,5-a]pyrimidin-7(4H)-one derivatives as Kv7.2 modulators. The pEC₅₀ is found to be significantly increased by the incorporation of –OH, –NO₂ or –Br groups in place of one –F, whereas –NH₂ has a negative effect on the pEC₅₀ values.

Lead nickel tungsten titanate Pb(Ni_1/3 Ti_1/3 W_1/3) O₃ single perovskite ceramics were prepared by using a high temperature solid-state reaction technique. The structural analysis of the sample confirms tetragonal crystal structure with space group P4mm having average crystallite size 73 nm, which is well supported from FTIR observations. The dielectric spectra with function of frequency and temperature suggest that sample is semiconducting in nature following NTCR behavior. The complex impedance results showed the contribution of grain and grain boundaries in the conduction mechanism. The activation energies were determined from the ac conductivity data in the temperature range of 200–350 °C and 360–500 °C. It has been observed that with the rise of frequency and temperature, the activation energy increases in the sample. The greater value of activation energy always supports the conduction mechanism due to hopping of the charge carriers. The semicircular arcs of Cole-Cole plots confirm that the sample is semiconducting in nature which supports our dielectric results.