Together with Pfizer Inc. and Digital Science, Oxford PharmaGenesis colleagues have co-developed an AI-based literature interrogation tool that will help researchers to identify biomarkers of emerging scientific interest in oncology.
Efforts to extract meaningful insights from publicly available scientific information are hampered by the limitations of standard keyword-based searching (such as that used when searching PubMed) and the sheer volume of available literature. Artificial intelligence (AI)-based approaches can help to overcome some of these limitations and allow researchers to glean insights that would otherwise remain buried in the scientific literature.
Dr Kim Wager and Dr Tom Rees from our Informatics and Data Science practice – together with partners Pfizer Inc. and Digital Science – have developed an AI-based tool that allows users to extract insights on biomarkers and biomarker panels from publicly available scientific information.
A data set comprising 726 cancer biomarkers was obtained from the Early Detection Research Network, an initiative of the National Cancer Institute. Using large-scale analytics of biomarker co-occurrence networks (i.e. instances of biomarker pairs being mentioned in close proximity in the same piece of text) based on full-text literature, the team successfully characterized biomarkers across six cancer types in terms of their emergence in the published literature and the context in which they are described.
Further work is now underway to optimize the utility of these biomarker co-occurrence networks for expert review and identification of potentially meaningful, emerging biologic relationships that can be applied in a clinical setting.
In April of this year, Dr Wager presented a poster detailing the development process and initial insights into the utility of the tool at the Annual Meeting of the American Association for Cancer Research.
Explore the interactive tool here.
For more information on our Informatics and Data Science capabilities, contact us here.
