Researchers at Edith Cowan University (ECU) have developed a groundbreaking method to measure biological age, which offers an innovative approach to detecting and monitoring age-related conditions. The study, titled Deep Reinforcement Learning–Driven Multi-Omics Integration for Constructing gtAge: A Novel Aging Clock from IgG N-glycome and Blood Transcriptome, was published in the journal Engineering on October 17, 2025.

Overview of the Research

The research team, which includes participants from Royal Prince Alfred Hospital in Sydney and Shantou University Medical College in China, explored blood elements that change with age. They particularly focused on the IgG N-glycome, referring to the sugar structures attached to antibodies, and a snapshot of gene activity within blood cells known as the transcriptome.

The gtAge Method

By integrating these two types of data through a sophisticated artificial intelligence (AI) technique known as Deep Reinforcement Learning, the researchers developed the gtAge aging clock. This new method predicts an individual’s age with an impressive 85% accuracy, which is significantly more precise compared to assessments made using IgG N-glycome or transcriptome data in isolation.

Furthermore, the study established that the variance between predicted age and actual age—referred to as delta age—is connected to critical health markers associated with aging, including cholesterol and blood sugar levels.

Understanding Biological Age

According to Dr. Xingang Li, a co-author and Postdoctoral Research Fellow at ECU’s School of Medical and Health Sciences, chronological age, which measures time since birth, does not fully reflect the complexities of the aging process. Dr. Li elucidates:

“In reality, some individuals remain healthy into their 80s and 90s, while others may experience age-related decline much earlier. This variation is due to differences in biological age, which factors in genetic, lifestyle, nutritional, and health influences to provide a more accurate depiction of the aging process.”

Dr. Li indicates that the gtAge framework accounts for 85.3% of the observed variance in chronological age.

A Cross-Disciplinary Approach

In this innovative study, the AI contributions were led by Dr. Syed Islam, Senior Lecturer in Computer Science at ECU. He described the development of a tailored AI tool named AlphaSnake, powered by Deep Reinforcement Learning:

  • Data Optimization: The algorithm intelligently selects the most relevant data points from both biological sources, avoiding the inefficiencies associated with indiscriminately merging data.

Testing the gtAge Method

The gtAge method was tested on 302 middle-aged adults participating in the Busselton Healthy Aging Study in Western Australia. Given Australia’s aging population, researchers believe that gtAge has the potential to be an invaluable tool in clinical settings.

Dr. Islam emphasizes the significance of measuring biological age rather than merely relying on an individual's birth date, stating:

“If we know in advance, we can adapt our lifestyle to better preserve our health and mitigate some of the wear and tear our bodies may experience.”

Future Directions

The research team, including Dr. Yao Xia, Dr. Syed Islam, Dr. Xingang Li, Dr. Abdul Baten, Dr. Xuerui Tan, and Professor Wei Wang, highlight that further research may enable better understanding and situations of age-related diseases. Continued developments could lead to:

  • Enhanced preventive health measures to extend vitality and lifespan.
  • Detailed insights into individual health risks based on biological age metrics.
  • Implementation of lifestyle interventions informed by biological aging assessments.

Conclusion

The gtAge method offers a novel lens through which we can view biological aging and its implications for health. As AI and multi-omics integration continue to evolve, the potential for improving individual health outcomes through better understanding of biological age becomes increasingly attainable.

References

1. Xia, Y., Islam, S., Li, X., Baten, A., Tan, X., & Wang, W. (2025). Deep Reinforcement Learning–Driven Multi-Omics Integration for Constructing gtAge: A Novel Aging Clock from IgG N-glycome and Blood Transcriptome. Engineering.

2. Retrieved from https://medicalxpress.com/news/2025-10-ai-powered-method-combines-blood.html