Scientists at the Icahn School of Medicine at Mount Sinai in New York have developed an artificial intelligence (AI) algorithm capable of predicting whether a drug could harm the developing fetus during pregnancy. The researchers have also created a “knowledge graph” tool that can identify existing medications not currently known to be harmful but could potentially cause congenital disabilities.
According to the scientists, birth defects occur in approximately one out of every 33 births in the US, and for most cases, the cause is unknown. Factors contributing to birth defects include genetic mutations as well as environmental influences like drugs, cosmetics, food, and pollutants that pregnant women may encounter.
The researchers trained their AI, called ReproTox-KG, using data from scientific literature on genetic associations, drug-induced gene expression changes, known drug targets, genetic burden scores, and the ability of small molecule drugs to cross the placenta.
“We wanted to improve our understanding of reproductive health and foetal development and, importantly, warn about the potential of new drugs to cause birth defects before these drugs are widely marketed and distributed. Although identifying the underlying causes is a complicated task, we offer hope that through complex data analysis like this we will be able, in some cases, to better predict, regulate, and protect against the significant harm that congenital disabilities could cause.”
– Avi Ma’ayan, director of the Mount Sinai Center for Bioinformatics and lead author of the paper
The AI employs a form of machine learning called semi-supervised learning (SSL), which uses a small set of labeled data to guide predictions for a much larger set of unlabeled data.
The AI identified over 30,000 preclinical small molecules that could potentially cause birth defects and identified more than 500 molecular mechanisms or “cliques” linking birth defects, genes, and drugs.
There have been instances in the past where a drug’s potential to cause congenital abnormalities went unnoticed during clinical testing and only became apparent after it had been approved and used by thousands of patients. One well-known example is the outbreak of limb malformations caused by thalidomide in the late 1950s and early 1960s. More recently, Sanofi’s epilepsy therapy Depakine (sodium valproate) has been linked to neurodevelopmental delays and conditions like spina bifida in children exposed to the drug in utero.
