Differentially expressed genes between the control cohort and the SARS-CoV-2 positive training set (A) Principal component analysis plot of the control cohort and the training set. Each data point represents an individual patient sample. (B) Volcano plot of differentially expressed genes between the control cohort and the training set. Each data point represents an individual gene. The genes selected by LASSO are those selected by LASSO as having predictive value. LASSO=least absolute removal and selection operator. Credit: The Lancet microbe (2024). DOI: 10.1016/S2666-5247(23)00363-4
Researchers at the University of Queensland have used machine learning to help predict the risk of secondary bacterial infections in hospitalized COVID-19 patients. The research is published in The Lancet microbe.
The machine learning technique can help detect whether antibiotic use is essential for patients with these infections.
Associate Professor Kirsty Short from the School of Chemistry and Molecular Biosciences said secondary bacterial infections can be extremely dangerous for people hospitalized with COVID-19.
“Estimates of incident secondary bacterial infections in COVID-19 patients are wide, but in some studies 100% of fatal cases experienced bacterial co-infection,” Dr. Short said.
“To reduce the risk of bacterial co-infections, it would theoretically be possible to treat all COVID-19 patients with antibiotics.
“However, there is a risk that overtreatment with antibiotics could lead to antibiotic resistance and the creation of superbugs.
“We helped develop a robust predictive model to determine the risk of bacterial infections in COVID-19 patients, facilitating prudent use of antibiotics. »
The technique is known as the “least absolute shrinkage and selection operator” – or LASSO for short.
Blood samples from COVID-19 patients from six countries were analyzed using the LASSO technique.
The team found that the expression of seven genes in a COVID-19 patient can predict their risk of developing a secondary respiratory bacterial infection after 24 hours of hospital admission.
Dr Meagan Carney from the School of Mathematics and Physics said these seven genes will now guide clinicians towards a more informed choice when it comes to antibiotic use.
“These data raise the exciting possibility that gene transcription and analysis at the time of clinical presentation in a hospital, along with machine learning, could be a game-changer in antibiotic prescribing,” said Dr. Carney.
She also pointed out that LASSO is simplified compared to the complex machine learning methods surrounding artificial intelligence that are currently being discussed in the media.
“Research projects like this, which use much less complex machine learning methods, can help bridge the gap between data scientists and scientists in other fields,” she said.
“We should strive to make data science less of a black box and inspire scientists around the world to better understand how it can revolutionize the medical industry.”
The researchers would like to acknowledge the extensive international collaboration of clinicians, virologists, bioinformaticians and many other experts who made this study possible, including the PREDICT consortium, the Snow Foundation and Nepean Hospital.
More information:
Meagan Carney et al, Host transcriptomics and machine learning for secondary bacterial infections in patients with COVID-19: a prospective, observational cohort study, The Lancet microbe (2024). DOI: 10.1016/S2666-5247(23)00363-4
Provided by the University of Queensland
Quote: Using machine learning to combat COVID-19 bacterial co-infection (February 2, 2024) retrieved February 2, 2024 from
This document is subject to copyright. Except for fair use for private study or research purposes, no part may be reproduced without written permission. The content is provided for information only.
