Astrazeneca Collaboration
More than 1 in 10 people worldwide are living with CKD.(1)
Up to 2 of 3 cases are caused by diabetes and high blood pressure.(2)
2-3% of the annual healthcare budget is spent on end-stage kidney disease treatment in high-income countries.(3)
Mortality due to CKD is increasing; rising from 0.9 to 1.2 million between 2005 and 2015.(4)
Our work
Chronic kidney disease (CKD) is an enormously complex disease with a high unmet need for new treatments. One of the most important decisions in the journey to finding a new treatment is selecting the right target. During this process, scientists must navigate complex CKD disease pathophysiology, many underlying mechanisms and vast amounts of data and information. Together with AstraZeneca, we are transforming the target discovery process by converting ‘big data’ into valuable knowledge, with the aim of getting more effective medicines to CKD patients.
The journey starts by building a CKD specific knowledge graph that leverages our vast Biomedical Knowledge Graph, in addition to CKD specific data, such as biological and chemical entities, their role and position within disease pathophysiology, and their relationships with other entities at a molecular, mechanistic, anatomic, and clinical level. Whereas data is usually analysed in silos, this holistic approach gives a full and unbiased overview of disparate datasets in order to draw connections between diseases and targets that may not be obvious from any single source.
After our drug discoverers have defined and refined the right biological questions, our powerful predictive target identification models reason across all of this information to uncover the patterns in the data and direct scientists towards high-quality novel targets that show the best promise as potential drug targets for CKD.
Next, our target triage user interface displays the biological rationale behind predicted targets, allowing the team of CKD experts to select the top candidates for testing. The AI-assisted target triage gives scientists confidence in their decision making and helps them make data-driven decisions over the best possible target to prioritise.
Not all CKD patients share the same underlying pathophysiology. Our precision medicine approach informs the process by helping to pinpoint the different molecular mechanisms driving the disease to identify subgroups that could respond similarly to a particular treatment. This helps identify the right patient population for the right target, which is fundamental for securing success in the clinic.
Atopic Dermatitis is the most common type of eczema and there is an urgent unmet need for an effective and safe alternative topical therapy for the broad patient population. To meet this need, we designed and developed a novel small-molecule treatment, BEN-2293, to address the debilitating itch and inflammation in both adults and children.
The cause of Idiopathic Pulmonary Fibrosis (IPF) is unknown, and the exact mechanisms involved in the pathogenesis of the disease are still unclear. Our collaboration with AstraZeneca aims to address this challenge by leveraging AI and machine learning to prioritise and select the right mechanism to target in the right patient population.
AstraZeneca has so far chosen to add the first novel AI-generated CKD target to enter their drug portfolio. We expect this will not be the last, and moving forward, we will continue to identify new targets with an embedded precision medicine strategy, using omics and clinical data to best understand how to treat different patient subgroups.
(1) Kidney Care UK. An estimated 1 in 10 people worldwide have chronic kidney disease. Available from: https://www.kidneycareuk.org/news-and-campaigns/news/estimated-1-10-people-worldwide-have-chronic-kidney-disease/ [Accessed January 2021].
(2) National Kidney Foundation. Kidney Disease: The Basics. Available from: https://www.kidney.org/news/newsroom/factsheets/KidneyDiseaseBasics. [Accessed January 2021].
(3) WHO. Global burden of kidney disease. Available at: https://www.who.int/bulletin/volumes/96/6/17-206441/en/ [Accessed January 2021].
(4) Wang H, et al. Lancet 2016; 288:1459544 https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(16)31012-1/fulltext
