A team of researchers have developed machine-learning software that can predict how dangerous a particular strain of Salmonella will be, according to a paper published in PLOS Genetics on Tuesday.
Salmonella is a nasty type of bacteria shaped like tiny jelly beans, and it causes stomach aches, diarrhea, nausea, vomiting, and all the other fun symptoms of food poisoning. The symptoms normally clear up after a few days if the bacteria only infects the gastrointestinal tract. But if it enters the bloodstream, it can lead to Typhoid fever and life-threatening septic shock.
The researchers therefore built a random forest classifier to differentiate between both types of Salmonella. First, they build a dataset collecting the DNA sequences from different types of Salmonella, including six strains that caused more severe infections, and seven gastrointestinal strains.
Next, they trained a random forest classifier to identify the 200 different genes that are more likely to give people food poisoning or Typhoid fever.
The random forest classifiers work by building multiple decision trees, where the nodes represent the different genes in the bacteria. The trees “predict a characteristic of the samples, in this case adaptation to an extraintestinal, or invasive, niche,” the boffins explained in their paper. After each Salmonella mutation is mapped out with the decision trees, the random forest classifier spits out an “impact score” that measures how invasive the strain is.
AI in Medicine? It's back to the future, Dr WatsonREAD MORE
"We have designed a new machine learning model that can identify which emerging strains of bacteria could be a public health concern. Using this tool, we can tackle massive data sets and get results in seconds,” said Nicole Wheeler, co-leader of the study and a postdoctoral fellow at the Wellcome Sanger Institute, a non-profit genomics and genetics research institute in the United Kingdom.
"Ultimately, this work will have a big impact on the surveillance of dangerous bacteria in a way we haven't been able to before, not only in hospital wards, but at a global scale."
The model can be applied to other bacterial problems like figuring out the genes responsible for antibiotic resistance for different bacterium.
“We are already using this approach to look for key differences in strains of Salmonella Typhi circulating in Asia compared to Africa,” said Nicholas Feasey, co-author of the paper and a researcher at the Liverpool School of Tropical Medicine.
"Instead of manually comparing the genomes of different strains of bacteria over weeks or months, we are able to discover the genetic changes behind emerging strains of bacteria in seconds. It offers the potential to study outbreaks in real time and thus rapidly inform public health strategies to control or prevent disease." ®