The Oxford Drug Discovery Institute is sorting through enormous volumes of biological data using artificial intelligence and "knowledge graphs," which might result in quicker therapies.

Artificial intelligence-powered databases are being used by researchers investigating Alzheimer's disease to speed up the medication development process by making it simpler to sort through enormous volumes of scientific data. Researchers at the Oxford Drug Discovery Institute in the United Kingdom claim that by employing those technologies, they can expedite the process of searching through journals and databases by almost ten times. This will enable them to more rapidly determine which genes or proteins should be chosen for additional research in order to produce possible Alzheimer's medications.

The Chief Scientific Officer of the Oxford Drug Discovery Institute, Emma Mead, stated that scientists have chosen 54 immune-related genes from a genome-wide association research, all of which are probably candidates for laboratory testing. These targets may include biological entities that prospective medications seek to influence, such as genes or proteins.

According to Mead, choosing Alzheimer's targets can be especially challenging because the illness has a large number of confusing environmental and socioeconomic risk factors, and there are several genes that might raise the chance of getting the condition.

However, in order for staff to more rapidly interpret those targets' properties across a wide range of sources—from the U.S. National Library of Medicine's PubMed to numerous scientific journals and its own datasets—it was necessary to use a knowledge graph, a database technology made popular by Google more than ten years ago for its search engine.

Relationships between individuals, concepts, and documents may be displayed using knowledge graphs, which are databases that depict information in a manner akin to maps. They have been employed in more recent years by sectors such as digital retail to provide online buyers with personalized recommendations.