The complex cell envelope of Gram-negative bacteria creates aformidable barrier to antibiotic influx. Reduced drug uptake impedes drugdevelopment and contributes to a wide range of drug-resistant bacterialinfections, including those caused by extremely resistant species prioritized bythe World Health Organization. To develop new and efficient treatments, abetter understanding of the molecular features governing Gram-negativepermeability is essential. Here, we present a data-driven approach, usingmatched molecular pair analysis and machine learning on minimal inhibitoryconcentration data from Gram-positive and Gram-negative bacteria to uncoverchemical features that influence Gram-negative bioactivity. Wefind recurringchemical moieties, of a wider range than previously known, that consistentlyimprove activity and suggest that this insight can be used to optimizecompounds for increased Gram-negative uptake. Ourfindings may help to expand the chemical space of broad-spectrum antibioticsand aid the search for new antibiotic compound classes.
展开▼
