Machine learning identifies drugs that could potentially help smokers quit — ScienceDaily

by admin
0 comment

Drugs like dextromethorphan, used to treat coughs caused by colds and flu, could be repurposed to help people quit smoking, according to a study by researchers at Pennsylvania State Medical College and the University of Minnesota. They said computer programs have developed new machine learning methods that analyze patterns and trends in data sets to identify drugs, some of which are already being tested in clinical trials.

Smoking is a risk factor for cardiovascular disease, cancer and respiratory disease, and kills nearly 500,000 people in the United States each year. Smoking behaviors may or may not be learned, but genetics also play a role in an individual’s risk for engaging in those behaviors. found that people with

Using genetic data from more than 1.3 million people, the study was jointly led by Dr. Dajiang Liu, Professor of Public Health Sciences, Biochemistry and Molecular Biology, and Bibo Jiang, Ph.D., Assistant Professor of Public Health Sciences. Large multicentre studies that use machine learning to study these large datasets – including specific data on individual genetics and self-reported smoking behavior.

Researchers have identified over 400 genes associated with smoking behavior. A person may have thousands of genes, so it was necessary to identify why some of those genes are associated with smoking behavior. Genes involved in signaling dopamine, the hormone that makes people feel relaxed and happy, made an easy connection. For the remaining genes, the research team had to determine the role each plays in biological pathways and use that information to identify drugs already approved to alter existing pathways.

Since most of the genetic data in this study are from European ancestry, the machine learning model not only studied that data, but also about 150,000 individuals with Asian, African, or American ancestry. It had to be adjusted to study smaller data sets as well.

Liu and Jiang worked with over 70 scientists on this project. They identified at least eight drugs that could be repurposed for smoking cessation, including dextromethorphan, commonly used to treat cold and flu coughs, and galantamine, used to treat Alzheimer’s disease. bottom. This research natural genetics Today, January 26th.

Liu, a researcher at the Pennsylvania State Cancer Institute and the Pennsylvania Huck Institute for Life Sciences, said: “Some of the drugs we identified are already being tested in clinical trials for their ability to help smokers quit, but there are still other candidates that could be explored in future research.”

Although machine learning methods have been able to incorporate small datasets from diverse ancestry, Jan said it remains important for researchers to build genetic databases from individuals with diverse ancestry.

“This will only improve the accuracy with which machine learning models identify individuals at risk of substance abuse and determine potential biological pathways targeted for useful treatment.”

Other authors from medical colleges participating in this project include Fang Chen, Xingyan Wang, Dylan Weissenkampen, Chachrit, Khunsriraksakul, Lina Yang, Renan Sauteraud, Olivia Marx, and Karine Moussa. They declare no conflict of interest.

This work was supported by the National Institutes of Health (grant grants R01HG008983, R56HG011035, R01HG011035, R56HG012358, R01GM126479, R21AI160138 and R03OD032630) and the Artificial Intelligence Program in Biomedical Informatics and Strategic Planning at the Pennsylvania State University of Medicine. The views of the authors do not necessarily represent those of the funders.

You may also like

Leave a Comment