Essential Oils as Antimicrobials against Acinetobacter baumannii: Experimental and Literature Data to Definite...

Area of Science:

• Natural Products Chemistry
• Computational Chemistry
• Microbiology

Background:

• Essential oils (EOs) possess broad biological activities but face clinical application challenges due to chemical variability and instability.
• Standardizing EOs by chemical composition is difficult; however, similar biological activities across different chemical profiles suggest standardization by effect is possible.
• Carbapenem-resistant *Acinetobacter baumannii* poses a significant threat, necessitating novel therapeutic strategies.

Purpose of the Study:

• To investigate the relationship between essential oil (EO) chemical composition and antibacterial activity against carbapenem-resistant *Acinetobacter baumannii*.
• To develop and validate machine learning models for predicting EO antibacterial efficacy based on chemical constituents.
• To explore the potential for designing biologically standardized EOs with optimized antibacterial and low cytotoxic properties.

Main Methods:

• Compiled a dataset of 82 EOs with known minimal inhibitory concentration (MIC) values from experimental data and the AI4EssOil database.
• Employed machine learning algorithms (SVM, Random Forest, Gradient Boosting, Decision Trees, KNN) to build quantitative composition-activity relationship (QCAR) models.
• Utilized Skater methodology for feature importance analysis and validated model predictions with experimental data and cytotoxicity assays.

Main Results:

• Machine learning models achieved 91% prediction accuracy for the antibacterial activity of new EO samples.
• Key chemical components like limonene, eucalyptol, alpha-pinene, and carvacrol were identified as critical for antibacterial efficacy.
• Analysis of cytotoxicity data revealed potential for designing EOs with a favorable balance of high antibacterial activity and low cytotoxicity.

Conclusions:

• Machine learning models can accurately predict essential oil antibacterial activity, facilitating the development of standardized EOs.
• Identification of key active components allows for the rational design and optimization of EOs for therapeutic applications.
• This approach supports the advancement of essential oils towards standardized, clinically applicable antimicrobial agents.