Abstract: BRIEF REPORT | BIOPHYSICS AND COMPUTATIONAL BIOLOGY OPEN ACCESS Computational prediction of interactions between Paxlovid and prescription drugs Yeji Kima,b,1 , Jae Yong Ryuc,1 , Hyun Uk Kima,b,d,e,1 , and Sang Yup Leea,b,e,2 Edited by Jens Nielsen, BioInnovation Institute, Copenhagen, Denmark; received December 29, 2022; accepted February 14, 2023 Pfizer’s Paxlovid has recently been approved for the emergency use authorization (EUA) from the US Food and Drug Administration (FDA) for the treatment of mild-to-moderate COVID-19. Drug interactions can be a serious medical problem for COVID-19 patients with underlying medical conditions, such as hypertension and diabetes, who have likely been taking other drugs. Here, we use deep learning to predict potential drug–drug interactions between Paxlovid components (nirmatrelvir and ritonavir) and 2,248 prescription drugs for treating various diseases. COVID-19 | drug interactions | DeepDDI2 | Paxlovid In December 2021, Pfizer’s Paxlovid (nirmatrelvir and ritonavir copackaged for oral use) received Emergency Use Authorization (EUA) from the US Food and Drug Administration (FDA) for the treatment of mild-to-moderate COVID-19 patients. Nirmatrelvir inhibits the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 3C-like protease to prevent virus replication, and ritonavir slows down the degradation of nirmatrelvir by acting as a CYP3A inhibitor. Subsequent clinical studies showed that Paxlovid is effective in reducing the hospitalization risk of COVID-19 patients aged 50 or older (1). A similar result was observed in patients aged 65 or older, who were treated with nirmatrelvir alone (2). Importantly, when the EUA was issued, FDA also provided information on 108 drugs that might exhibit potential drug interactions with Paxlovid (3). Likewise, in January 2022, European Medicines Agency (EMA) also reported 128 drugs that can potentially interact with Paxlovid (‘Paxlovid: EPAR—Product information’ available at: https://www. ema.europa.eu/en/medicines/human/EPAR/paxlovid#product-information-section; Accessed December 22, 2022); FDA and EMA reported 69 drugs in common. Both FDA and EMA reported potentially interacting drugs mainly because these drugs can serve as substrates or inhibitors of cytochromes P450 (CYPs), leading to unwanted drug–drug interactions (DDIs). Such DDIs can be a serious problem for the COVID-19 patients having underlying medical conditions such as hypertension and diabetes because these patients are already taking medicine to treat their conditions (4–7). A problem here is that there are likely more drugs that might interact with Paxlovid, and possible DDIs involving Paxlovid cannot be experimentally examined in a short period of time. Here, we report the list of a large number of prescription drugs that are predicted to have DDIs with Paxlovid by employing DeepDDI. DeepDDI is a computational model developed using deep learning that predicts the pharmacological effects and adverse drug events (ADEs) of DDIs (8). DeepDDI receives structural information as simplified molecular-input line-entry system of two drugs in a pair as an input, and predicts DDI types as an output in the form of human-readable sentences. The DeepDDI output sentences describe changes in the pharmacological effects and/or the risk of ADEs as a result of the DDI. DeepDDI originally..