Related Experiment Videos
Prediction of Antibiotic Interactions Using Descriptors Derived from Molecular Structure
Daniel J Mason1, Ian Stott2, Stephanie Ashenden1
1Centre for Molecular Informatics, Department of Chemistry, University of Cambridge , Cambridge CB2 1EW, United Kingdom.
Journal of Medicinal Chemistry
|April 7, 2017
Summary
This study introduces a computational method to predict effective antibiotic combinations using drug molecular structures. The approach successfully identified new synergistic antibiotic pairs, aiding the fight against bacterial infections.
Area of Science:
- Computational biology
- Medicinal chemistry
- Pharmacology
Background:
- Combination antibiotic therapies are crucial for combating bacterial infections.
- Identifying effective antibiotic combinations is challenging due to the vast search space.
Purpose of the Study:
- To develop a computational framework for predicting antibiotic interactions and synergy.
- To validate the framework's predictive accuracy using experimental data.
Main Methods:
- Utilized substructure profiles derived from drug molecular structures.
- Developed a predictive model for antibiotic synergy.
- Experimentally validated predictions on new drug pairs.
Main Results:
- Substructure profiles effectively predicted synergy in a dataset of 153 drug pairs.
- Identified 37 new synergistic antibiotic pairs from 123 experimentally tested pairs.
- Achieved a 2.8-fold enrichment in predicted synergistic pairs, with 10 out of 12 validated.
Conclusions:
- The computational framework reliably predicts antibiotic interactions based solely on chemical structures.
- This methodology can accelerate the discovery of novel synergistic antibiotic combinations.
- The approach is applicable to any antibiotic pair within the model's domain.