Korean researchers develop AI to foretell hostile drug reactions with Pfizer’s Paxlovid


A analysis crew from the Korea Superior Institute of Science and Know-how has developed an AI mannequin to foretell hostile reactions between oral anti-COVID-19 treatment and pharmaceuticals.

Researchers from KAIST’s Division of Biochemical Engineering made a brand new model of the DeepDDI AI-based drug interplay prediction mannequin to verify how ritonavir and nirmatrelvir, two elements of Paxlovid by pharmaceutical large Pfizer, would work together with pharmaceuticals. 

The brand new mannequin DeepDDI2 can compute for and course of a complete of 113 drug-drug interplay sorts, a press launch famous.

It was later discovered that Paxlovid interacts with roughly 2,248 pharmaceuticals: 1,403 medicine with ritonavir and 673 medicine with nirmatrelvir. 

The researchers then proposed various choices for pharmaceuticals with excessive hostile reactions with Paxlovid: they discovered 124 medicine with low potential hostile reactions with ritonavir and 239 medicine with nirmatrelvir.


COVID-19 sufferers with comorbidities, equivalent to hypertension and diabetes, are prone to be taking antiviral treatment with different medicine. Nonetheless, drug-drug interactions and hostile drug reactions with Paxlovid “haven’t been sufficiently analysed,” the KAIST researchers stated. Utilising AI expertise, they then got down to discover how the continued use of antiviral remedy with different medicine might result in severe and undesirable issues. 


Pfizer is inching near getting the US Meals and Drug Administration’s full approval for Paxlovid. This comes as an advisory panel final week voted to suggest the approval because it deems the drug protected and efficient. The corporate acquired emergency use approval for Paxlovid from the regulatory physique in December 2021. Following the advisers’ vote, it’s anticipated that the US FDA will make a last resolution on its full approval by Might. 


“The outcomes of this research are significant at occasions like once we must resort to utilizing medicine which are developed in a rush within the face of pressing conditions just like the COVID-19 pandemic. [With DeepDDI2], it’s now doable to establish and take mandatory actions towards hostile drug reactions brought on by drug-drug interactions in a short time,” KAIST Professor Sang Yup Lee stated in a press release.

About Author

Leave a Reply

Your email address will not be published. Required fields are marked *