Optimization of Doses of Antibiotics and Cuminaldehyde to Combat Methicillin-Resistant Staphylococcus aureus (MRSA): A Study With Machine Learning.

Methicillin-resistant Staphylococcus aureus (MRSA), a drug-resistant organism, can cause a spectrum of infections in the human host involving biofilm. Therefore, novel therapeutic approaches need to be explored to mitigate this persistent infection. This study investigated a combinatorial approach that incorporates cuminaldehyde (a phytochemical) alongside aminoglycoside antibiotics (gentamicin and tobramycin) to improve antibiofilm efficacy by addressing multiple targets. In this regard, to achieve precise dosing of the chosen compounds for effective biofilm management, different machine learning models, namely multiple linear regression (MLR), polynomial regression (PR), artificial neural network regression (ANNR), and support vector regression (SVR), were employed. The results suggested that ANNR exhibited a strong association between the predicted and experimental observations (R2 = 98.07). Furthermore, the ANNR model, followed by genetic algorithm (GA), recommended that the combinatorial doses of the selected compounds [cuminaldehyde (40 μg/mL); gentamicin (0.5 μg/mL); and tobramycin (0.035 μg/mL)] could show the highest antibiofilm activity against MRSA. Additionally, this study revealed that the combination of the mentioned compounds at their recommended doses not only accumulated intracellular reactive oxygen species (ROS) but also increased the cell membrane permeability of MRSA. Thus, this study provides a promising foundation for developing novel therapeutic strategies against MRSA biofilm through an AI-driven approach.