A metagenomics-based tool for surveillance of bacterial pathogens in environmental samples

Abstract

Bacterial pathogens are responsible for millions of human deaths and significant losses in food production annually. Soil acts as a natural reservoir for many critical bacterial pathogens, however, the lack of accurate, high-throughput methods to identify the distribution of bacterial pathogens in soil and other environments greatly limits our ability to develop effective surveillance and disease management strategies and policies. Here, we present MetaPatho, an easy-to-use, rapid, robust, and shotgun metagenomics-based tool for identifying human and plant bacterial pathogens in environmental metagenomes. We first constructed high-quality pathogenic databases for the identification of human and plant bacterial pathogens. In-silico tests and soil spike-in experiments, including pathogen DNA spike-in and cell spike-in tests, were then employed to examine the accuracy and robustness of our tool. The in-silico tests suggested that the tool accurately identified pathogenic sequences, achieving an average accuracy of 97.4%, with a specificity of 98.1% and a sensitivity of 97.2%. Further analyses from soil spike-in experiments showed that the tool accurately detected bacterial pathogens at different concentrations, achieving an average accuracy of 96.3%. Overall, our tools can provide a high-throughput and accurate surveillance of bacterial pathogens in environmental metagenomes, contributing to an effective risk management associated with human and plant diseases.