The group of bacteria known as Salmonella includes many different types that vary in the severity of the disease they cause. Some types cause food poisoning, whereas others can cause invasive life-threatening disease (e.g. typhoid fever). A reproducible contellation of genomic changes occurs when Salmonella adapt to an invasive lifestyle, which can be detected using a machine learning algorithm.

​

This algorithm can be used to flag emerging invasive strains. In the past, we've only noticed these dangerous strains once they have already infected large numbers of patients, and when we trace their origins we find they emerged and started spreading around the world decades earlier but we failed to notice them. This software lets us identify these strains as they appear, which could help us prevent their spread.

​

The algorithm has already been used to characterize a new invasive strain of Salmonella Typhimurium found in the Democratic Republic of the Congo, and to better understand the impact of human activity on the evolution of Salmonella Typhimurium. The model is available on Github, with a guide to help you analyse your samples. Alternatively, you can email me and I may be able to analyse them for you. 

​