Random Forest model resolves the challenges against Multi-Drug Resistant bacteria by AgNPs

The crisis against Multi-Drug Resistant (MDR) bacteria is on the rise globally. The attempts to discover new solutions like the invention of new antibiotics have encountered frequent failures due to the relatively high speed of microbial resistance. Among the hopeless tries, nanotechnology has introduced unique materials that can function as promising anti-MDR agents. Nanoparticles (NPs) especially silver nanoparticles (AgNPs) and their corresponding derivatives have been among the most outstanding antimicrobial and anti-MDRs due to their magnificent physical and chemical properties. Due to high costs and numerous tries and errors, there should be some strategies for predicting the anti-MDR activities of NPs. In the present study, a Machine Learning (ML) based model; Random Forest (RF) was applied to predict the anti-MDR activities of AgNPs. Once, the literature was provided, the desired information regarding the physical and chemical information besides the taxonomical information of the MDR bacteria was retrieved. Then, the preprocessing strategies were applied. Subsequently, the model was predicted with a high accuracy (R²=0.73). The analysis of significant attributes revealed that Dose, DLS_size, Core_size, and species are the most important factors in the anti-MDR activities of AgNPs. The results proved this tool can help scientists to have reasonable assumptions toward anti-MDR activities of AgNPs before any experiments, cutting the high costs of numerous experiments.