Understanding how different drugs can interact with one another is vital for patient wellbeing. While some drug combinations can cause worsening symptoms or dangerous complications, others may alleviate symptoms and help the patient to recover more quickly. Therefore, knowing which drug combinations to pursue and which to avoid is critical. In many cases, the effects of drug combinations arise because two drugs bind to the same target protein, particularly if they share similar properties. However, this is not the case for all drug interactions. Read More
Historically, drug interaction studies have involved expensive and time-consuming clinical and laboratory-based experiments. This research has mostly focused on shared protein binding sites or drug-gene interactions. However, protein-linking networks are present in our cells, meaning that drugs can still interact with one another even if they don’t bind to the same protein.
To investigate this further, Dr. Jennifer Wilson at UCLA and her colleagues developed a new computational model to predict drug interactions. The team’s model considers drugs that do not share protein binding sites, but whose binding sites are linked via the cellular network.
To begin with, the researchers first collected adverse drug reactions associated with single drugs by searching drug labelling records and extracting them using a computer algorithm.
Next, they identified links between the target proteins of these drugs and proteins known to be associated with adverse effects. They grouped drugs that shared protein links and adverse effects to form a new classification system. Drug-drug interactions were then predicted according to their newly attributed classification and validated using clinical data, to estimate the potential effect of each drug combination.
Using their new modelling approach, Dr. Wilson and her team created an extensive list of proteins that are associated with adverse drug reactions, many of which were previously unknown. Analysis of cellular networks revealed fresh associations between various drugs and certain conditions.
The researchers also accurately detected rare drug-drug interactions that would have ordinarily been overlooked. In addition, they discovered previously obscure protective effects exerted by specific drug combinations.
The team’s model provides an accessible and straightforward approach to predict the effects of drug combinations in treating various conditions, and to aid the process of drug development. Considering drugs that bind to connected proteins in addition to drugs with shared binding proteins is an effective way to predict adverse interactions and their impact on patient safety and disease outcomes.
The team’s work will help medical professionals to avoid causing harm by unintentionally mixing incompatible drugs, while also enabling them to improve patient outcomes by harnessing beneficial combinations.
