Essential oils (EOs) are vastly used as natural antibiotics in Complementary and Alternative Medicine (CAM). Their intrinsic\nchemical variability and synergisms/antagonisms between its components make difficult to ensure consistent effects through\ndifferent batches. Our aimis to evaluate the use of artificial neural networks (ANNs) for the prediction of their antimicrobial activity.\nMethods. The chemical composition and antimicrobial activity of 49 EOs, extracts, and/or fractions was extracted from NCCLS\ncompliant works. The fast artificial neural networks (FANN) software was used and the output data reflected the antimicrobial\nactivity of these EOs against four common pathogens: Staphylococcus aureus, Escherichia coli, Candida albicans, and Clostridium\nperfringens as measured by standardised disk diffusion assays. Results. ANNs were able to predict >70% of the antimicrobial\nactivities within a 10mm maximum error range. Similarly, ANNs were able to predict 2 or 3 different bioactivities at the same\ntime.The accuracy of the prediction was only limited by the inherent errors of the popular antimicrobial disk susceptibility test and\nthe nature of the pathogens. Conclusions. ANNs can be reliable, fast, and cheap tools for the prediction of the antimicrobial activity\nof EOs thus improving their use in CAM.
Loading....