Approach has potential to transform early disease detection and drug discovery
Research published today in Nature Genetics introduces MILTON, a cutting-edge machine-learning research tool with the ability to predict more than 1,000 diseases before diagnosis. Drawing on comprehensive data from nearly 500,000 participants in UK Biobank*, MILTON assesses and can identify whether individuals are more likely to have and be diagnosed with certain diseases. This breakthrough will accelerate the discovery of new drug targets and biomarkers paving the way for the development of more effective and targeted treatments.
MILTON, which stands for MachIne Learning with phenoType associatiONs, is designed to identify and predict disease with high accuracy, having been developed using data from 67 routine clinical biomarkers. It also incorporates data from 3,000 plasma protein measurements for a subset of 50,000 UK Biobank participants. Using the area under the curve (AUC) metric**, a gold standard for evaluating machine learning performance, MILTON analysed 3,200 diseases, achieving a high predictability level for 1,091 diseases (AUC above 0.7) and exceptional performance for 121 diseases (AUC above 0.9). These findings reveal potential new biomarker signatures that can predict disease risk years before diagnosis.
MILTON's capabilities are set to enhance drug target discovery in large-scale, population-based genetic studies. In conventional case-control studies, hospital billing codes and self-reported data often classify participants into cases and controls, which can be incomplete. However, MILTON enables users to identify individuals who may have been incorrectly classified as controls. This AI-augmented reclassification was found to significantly enhance the statistical power for genetic discovery, expanding the scope and accuracy of gene discovery for hundreds of diseases.
Slavé Petrovski, Vice President, Centre for Genomics Research, AstraZeneca said, “MILTON is a significant advance on the predictive tools currently used. Our research demonstrates MILTON’s capabilities and how it is able to identify disease risk cases in large biobank datasets, which in the future, could enable us to detect illnesses earlier and at more treatable stages. Improving our ability to detect illnesses earlier and at more treatable stages is critical for early interventions in clinical care.”
Euan Ashley, Associate Dean, Professor of Medicine, Genetics & Biomedical Data Science, Stanford School of Medicine & co-author of the study said, “MILTON goes beyond the genetic code to consider the breadth of molecular factors that are associated with the risk of developing disease. This AI technology outperforms traditional diagnostic methods and polygenic risk scores and offers significant potential to transform drug discovery and accelerate the future of preventative medicine.”
MILTON's predictive capabilities, covering over 1,000 diseases, can be applied to any biobank irrespective of genomic ancestry. As AstraZeneca continues to expand its capabilities, including adding additional proteomic data, MILTON's ability to predict a wider range of diseases will improve.
As part of AstraZeneca’s commitment to open science, all gene-disease associations and predictive biomarker collections are available to the public through an interactive portal (http://milton.public.cgr.astrazeneca.com).
Notes
AstraZeneca’s Centre for Genomics Research
The Centre for Genomics Research (CGR) is a unified and centralised multi-omics and target-indication discovery unit within AstraZeneca. Working in close partnership across therapy areas, the CGR comprises four core functions in genomic sciences: Human Genetics (AstraZeneca’s Genomics Initiative), Dynamic Omics, Functional Genomics, and Translational Genomics. In partnership with Therapy Area colleagues, the Centre’s patient-centric focus supports the development of robust therapeutic strategies anchored in causal human biology.
AstraZeneca
AstraZeneca (LSE/STO/Nasdaq: AZN) is a global, science-led biopharmaceutical company that focuses on the discovery, development, and commercialisation of prescription medicines in Oncology, Rare Diseases, and BioPharmaceuticals, including Cardiovascular, Renal & Metabolism, and Respiratory & Immunology. Based in Cambridge, UK, AstraZeneca’s innovative medicines are sold in more than 125 countries and used by millions of patients worldwide. Please visit astrazeneca.com and follow the Company on social media @AstraZeneca.
Contacts
For details on how to contact the Investor Relations Team, please click here. For Media contacts, click here.
References
*UK Biobank is the world’s most comprehensive source of health data available for research, housing a vast and continuously growing dataset of biological, health and lifestyle information collected over 15 years from half a million UK volunteers.
**AUC ranges in value from 0 to 1. A model where predictions are 100% wrong has an AUC of 0.0; one where predictions are 100% correct has an AUC of 1.0.