PATHOGEN ADAPTATION WITHIN THE HUMAN HOST

Microbial infections, and the rising resistance to antibiotics in particular, are a major public health concern. Current pathogen diagnostic is based on culturing of clinical samples. This approach is inherently slow and, more importantly, does not allow prediction of the potential for evolution of antibiotic resistance. We hypothesize that the current resistance profile of a bacterium as well as some of its adaptive potential are encoded in its genome. Therefore, we suggest the use of whole-genome sequencing together with machine-learning based algorithms as an alternative diagnostic approach which will allow us to identify (1) the type of the pathogen (2) resistance profile and (3) resistance potential. As a first step, we will use Staphylococcus aureus (S. aureus) isolates as a case study to reveal genomic determinants of pathogen resistance. S. aureus, especially methicillin- resistant S. aureus (MRSA), is one of the most important human pathogens, causing various infections, including life threatening infections. Its increasing resistance to antibiotics complicates treatment and poses a public health challenge.

Through a collaboration between our lab and the Government Central Laboratories (GCL) of the Israeli Ministry of Health, we have obtained a large collection of S. aureus isolates. In this collection we will focus on patients with multiple isolates from different time points. We will sequence and identify the differences among same- patient isolates, to infer on genetic determinants of pathogen resistance. We will use our previously developed pipeline to identify SNPs and short indels separating strains as well as gene content of each isolate. To deduce the probability of a specific isolate to evolve resistance in the clinic we will measure how many mutations were required for a sensitive strain to become resistant.

In conclusion, we aim to develop and apply predictive diagnostics for the pathogen’s current antibiotic resistance as well as the potential for evolution. We expect this to be a step toward a sensitive and accurate genome-based diagnosis in clinical practice.