A University of Otago expert is a member of an international partnership that has developed a novel machine learning tool that can identify whether budding strains of the Salmonella bacterium are more prone to result in severe bloodstream infections instead of food poisoning only.
The software tool, developed by a researcher at the Wellcome Sanger Institute in the UK, the Helmholtz Institute for RNA-based Infection Research in Germany, and the University of Otago, greatly accelerates the procedure for discovering the genetic alterations underlying the new invasive varieties of Salmonella that are a huge threat to public health.
The bacterial group called Salmonella consists of several diverse types that differ in the severity of the infection they bring about. Food poisoning is caused by few strains, called Gastrointestinal Salmonella, whereas severe disease is caused by other strains by thinning out further than the gut, for instance, Salmonella typhi that causes typhoid fever.
The technique accurately categorizes which Salmonella types will be aggressive and which ones aren’t, by examining the genome sequences, as said by Dr Paul Gardner of the Department of Biochemistry of the University of Otago.
To comprehend the genetic alterations that find out whether a budding Salmonella strain will lead to food poisoning vs. a more serious infection, the team designed the machine learning model that examines which mutations play a significant part.
The team skilled the model making use of old salmonella lineages that are evolutionarily discrete, comprising 6 Salmonella bacteria that led to invasive infections and & gastrointestinal strains. The machine learning model recognized nearly 200 genes entailed in finding out whether the bacterium will lead to food poisoning or is modified to cause invasive infection.
A few days back, the Government invited the Artificial intelligence professionals from the University of Otago to work with them to shape an AI and predictive analytics framework.