Abstract

Carbapenems are β- lactam antibiotics used to fight infections caused by organisms characterized by multidrug resistance. The use of β-lactams for over 60 years has led to a dramatic increase in resistance, which resulted in a decrease in the effectiveness of many antibiotics in this group. European Centre for Disease Prevention and Control successively informs about the emerging resistance to carbapenems. Aim of the present study was to determine the relationship between physicochemical parameters and the MIC₅₀ values of meropenem, imipenem, doripenem, tebipenem, and ertapenem designated for Streptococcus spp., Klebsiella spp., Haemophilus spp., Pseudomonas spp., Staphylococcus spp., and Proteus mirabilis. Quick- Prop 3.1 software from Schrödinger package v 31207 was used for calculations. The leave-one-out method was used for model cross-validation. Squared cross-validated correlation coefficient (Q²) parameter and differences between Q² and R² were calculated as measure of the internal performance and model predictive ability. Difference of ability between fitting and predictive ability was analyzed using difference between asymptotic squared cross-validated correlation coefficient (Q² _asym) and Q². In the presented work, it has been shown that it is possible to correlate physicochemical parameters groups (in the form of arithmetic expressions) with MIC₅₀ values. The results of the analysis lead to a hypothesis that, depending on the differences between strains of bacteria various physicochemical parameters of arithmetic expressions reflect the diversity of interactions at the molecular level, despite the common mechanism of action. The presented models illustrate the correlations of sum of interactions between drug molecule and bacteria, expressed by MIC₅₀ value.

Keywords: Carbapenems, in silico, MIC, structure, QSAR, validation.

Graphical Abstract