Cancer management relies on the prolonged use of potent chemotherapy regimens. However, these drug cocktails can trigger drug-drug interactions (DDIs), and about 20% of all cancer patients experience at least one clinically relevant DDI. Given the complexity of biological systems and the inability to recapitulate disease context, experimental models for accurately predicting DDIs have limited success. Hence, in silico methods for DDI prediction have become increasingly popular in the BioPharma industry in the past decade.

One of the popular in silico methods for big data mining is Knowledge Graphs (KG).KGs use powerful data analytics techniques that can assimilate and model large volumes of multi-faceted data to extract meaningful information in a short time. KG-based DDI prediction models can have an impact on multiple stakeholders in the oncology ecosystem and hold significant promise in predicting the side effects of chemotherapy cocktails, thereby improving overall cancer care outcomes.

For the past decade, Persistent has been at the forefront of applying in silico technologies to support Bio/Pharma R&D pipelines. We have recently demonstrated the use of KGs for the identification of potential DDI from a commonly used lung cancer chemotherapy cocktail. It is a must-read whitepaper on the application of knowledge graphs for drug discovery and the future scope for big data mining for cancer chemotherapy.