Building up of the predictive quantitative structure-property/activity relationships (QSPRs/QSARs) for nanomaterials usually are impossible owing to the complexity of the molecular architecture of the nanomaterials. Simplified molecular input-line entry system (SMILES) is a tool to represent the molecular architecture of "traditional" molecules for "traditional" QSPR/QSAR. The quasi-SMILES is a tool to represent features (conditions and circumstances), which accompany the behavior of nanomaterials. Having, the training set and validation set, so-called quantitative feature-property relationships (QFPRs), based on the quasi-SMILES, one can build up model for zeta potentials of metal oxide nanoparticles for situations characterized by different features.

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 rm(2), r(*)m(2), [Formula: see text], 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.

: Today, the scientific community across the globe generates huge data; for example more than 74 million substances are registered in Chemical Abstract Services. Another estimate tells that there is about 1060 molecules classified as new drug like molecules. This huge space is now refereed as ‘dark chemical space’ or ‘dark chemistry’. Today one can see a surge in the number live databases (protein, cell, tissues, structure, drugs etc.) and every day these are updated with new information. So, the synchronization of the three different sciences ‘genomics’, proteomics’ and ‘in-silico simulation’ is essential and it will revolutionize the process of drug discovery. Now, the screening of the large number of drug like molecules is a challenge and it should be done in an efficient manner. Virtual screening (VS) is an important computational tool in the drug discovery process, however the drugs are need to be verified experimentally at every stage. Amongst, the various VS methods, quantitative structure–activity relationship (QSAR) analysis is the proven and accepted machine learning technique. QSAR is well-known for its high and fast throughput screening with good hit rate. The QSAR model building involves, (i) chemogenomics data collection from a database or literature, (ii) Calculation of right descriptors from molecular representation, (iii) establishing a relationship (model) between biological activity & the selected descriptors (iv) application of QSAR model to predict the biological property for the molecules. Therefore, it is desirable to test experimentally the hits obtained by VS. In this mini-review: different web-based machine learning tools, techniques of QSAR for VS, successful applications of QSAR based VS leading to the drug discovery and advantages and challenges of this powerful technique were highlighted.

No abstract