QSAR analysis of poliovirus inhibition by dual combinations of antivirals

• Published in: Structural chemistry Vol. 24; no. 5; pp. 1665 - 1679
• Main Authors: Muratov, E. N., Varlamova, E. V., Artemenko, A. G., Polishchuk, P. G., Nikolaeva-Glomb, L., Galabov, A. S., Kuz’min, V. E.
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01-10-2013; Springer Nature B.V
• Subjects: Binary mixtures; Chemistry; Chemistry and Materials Science; Computer Applications in Chemistry; Molecular structure; Original Research; Physical Chemistry; Picornavirus; Poliovirus 1; Replication; Theoretical and Computational Chemistry; SiRMS; Mixture descriptors; Poliovirus; Dual combinations
• ISSN: 1040-0400; 1572-9001
• DOI: 10.1007/s11224-012-0195-8

We have applied Hierarchical QSAR Technology (HiT QSAR) to the prediction of antiviral effects of paired combinations of picornavirus replication inhibitors against poliovirus 1 (Mahoney) in vitro. The inhibition from all binary combinations of eight antivirals were investigated. Simplex representation of molecular structure (SiRMS) was used for the generation of molecular descriptors of both pure compounds and all dual mixture combinations. Predictive QSAR models were obtained using the partial least squares (PLS) method. Predictive power of the developed models was validated using eightfold external cross-validation (CV,
 = 0.67–0.93). Adequate models (
 = 0.53–0.97) were obtained in the same way for predicting measured inhibitory concentrations at other levels (i.e., IC
30
, IC
40
, IC
60
, IC
70
). The usage of predicted values of these concentrations in the framework of the feature net (FN) approach led to an insignificant increase in the quality of the obtained QSAR models (
 = 0.71–0.94). Developed QSAR models were analyzed and interpreted so that structural fragments and components of the combination promoting the antiviral activity were determined (e.g., 2-(4-methoxyphenyl)-4,5-dihydrooxazole or the combination of
N
-hydroxybenzimidoyl and 3-methylisoxasole). Then the resulting consensus model was used to predict novel potent combinations of drugs. Combinations of enviroxime with pleconaril, WIN52084, and rupintrivir and the mixture of rupintrivir with disoxaril were predicted to cause the most inhibition of poliovirus 1 replication. HiT QSAR proved itself as an adequate tool for QSAR analysis of mixtures and, although the method described here is suitable only for binary mixtures, it can be easily extended for more complex combinations.