Related Experiment Video
Updated: Dec 20, 2025
11:34
High-throughput Screening for Broad-spectrum Chemical Inhibitors of RNA Viruses
Published on: May 5, 2014
14.2K
Experimental Data Based Machine Learning Classification Models with Predictive Ability to Select in Vitro Active
Manuela Sabatino1, Marco Fabiani2, Mijat Božović3
1Rome Center for Molecular Design, Department of Drug Chemistry and Technology, Sapienza University, P.le Aldo Moro 5, 00185 Rome, Italy.
Molecules (Basel, Switzerland)
|May 30, 2020
Summary
Machine learning models effectively predicted essential oils' antiviral activity against HSV-1. This approach identified potent and low-toxicity natural remedies, including Calamintha nepeta oil, for potential therapeutic use.
Area of Science:
- Natural Product Chemistry
- Computational Biology
- Virology
Background:
- Essential oils are increasingly studied for therapeutic properties, including antiviral applications.
- Herpes Simplex Virus type 1 (HSV-1) remains a significant viral pathogen.
- Predictive modeling offers a novel approach to identify effective natural compounds.
Purpose of the Study:
- To investigate an in-house library of essential oils for potential inhibition of HSV-1 infection.
- To develop and validate machine learning models for predicting essential oil bioactivity and toxicity.
- To identify highly active and low-toxicity essential oils for further study.
Main Methods:
- In vitro testing of essential oils against HSV-1 infection, determining IC50 and CC50 values.
- Gas-chromatography/mass spectrometry for chemical analysis of essential oil composition.
- Machine learning classification models (Partial Least Square Discriminant Analysis) trained on experimental data.
Main Results:
- Machine learning models successfully predicted the antiviral activity and toxicity of untested essential oils.
- Four out of five selected essential oils demonstrated high activity and low toxicity.
- Calamintha nepeta oil (CJM1) exhibited the highest potency and selectivity index (SI > 47.5).
Conclusions:
- Machine learning provides a valuable tool for predicting the bioactivity of complex natural mixtures like essential oils.
- This multidisciplinary approach can accelerate the discovery of effective antiviral agents.
- Future research can focus on designing optimized essential oil blends with enhanced potency and reduced toxicity.
Related Concept Videos
01:28Antimicrobial Effectiveness
841
The effectiveness of antimicrobial agents depends on various factors influencing their ability to eliminate microbial populations. Larger microbial populations require more time for complete eradication, emphasizing the importance of population size analysis when evaluating antimicrobial efficacy.Microbial resistance to antimicrobial agents varies significantly. Highly resilient microorganisms include endospores, gram-negative bacteria, and non-enveloped viruses, while prions are exceptionally...
841
01:27Chemical Agents for Microbial Control
652
Chemicals play important roles in controlling microbial growth by targeting microbial structures and functions as sanitizers, antiseptics, disinfectants, and sterilants.Alcohols are commonly used sanitizers, effectively disrupting lipid membranes, which compromises cell integrity. They are also used as antiseptics and disinfectants due to their rapid action and versatility.Phenols and their derivatives phenolics , known for denaturing proteins and disrupting cell membranes, are particularly...
652