Finding new sources of antibiotics using phylogenetic, spatial, and phytochemical data

Just 90 years after the discovery of penicillin, the world may soon be facing a post-antibiotic era, as emerging multidrug-resistant bacteria are threatening global public health. Plants have been used medicinally for thousands of years, but only a fraction have been evaluated for biological activity.

To facilitate the discovery of novel plant compounds with potential as antibiotic agents, this study classified more than 16,000 metabolites according to bioactive properties from more than 7,500 seed plant species in Java, Indonesia, one of the world's megadiverse countries. They then searched for phylogenetic signals of anti-infective properties, identifying "hot" clades of overrepresentation and "cold" clades of underrepresentation.

Using GBIF-mediated occurrences of Javanese seed plants, the authors then constructed distribution models to assess the spatial patterns of plants with anti-infective properties in relation to total plant diversity on the island.

Their phylogenetic analysis identified 26 "hot" clades, including 25 whole families, e.g. daisies/sunflowers (Asteraceae), and 24 "cold" clades, including 16 entire families, e.g. grasses (Poaceae) and orchids (Orchidaceae). The spatial analysis showed that regions with the highest diversity correlated with the highest probability of finding novel antibiotic compounds.