Biomarkers of Aging Challenge Series

Stimulating innovation and collaboration in the development of next-generation aging biomarkers.

Made possible and supported by:

The Challenge

Aims to generate the most accurate predictive models for aging biomarkers, including mortality and multi-morbidity.

In addition, the Challenge fosters technological advancements for large-scale profiling and reduces costs for biomarker development and validation.

The competition leveraged a high-quality dataset featuring proteomic profiles, DNA methylation data, and health outcomes from over 500 individuals. These datasets allow us to develop predictive models for chronological age, mortality risk, and future disease incidence.

Three Consecutive Phases

  1. Chronological Age Prediction Challenge
    Phase 1 kicked off in Q1 2024 with $30,000 in awards. This phase completed in Q3 2024.

  2. Longevity Prediction Challenge (Mortality)
    Phase 2 kicked off in Q3 2024 with $70,000 in awards. This phase completed in Q4 2024 and winners were celebrated at the 2024 Biomarkers of Aging Conference.

  3. Healthspan Prediction (Multi-Morbidity)
    Phase 3 will kick off in Q3 2025 with
    $100,000 in awards!

Congratulations to our Phase 1 and Phase 2 Winners

Phase 1 - Chronological Age

Phase I leveraged a unique, high-quality dataset that includes DNA methylation and aging outcome data for over 500 diverse individuals. DNA methylation and other first-generation biomarkers of aging are often trained to predict chronological age. Deviations between predicted and actual age (prediction errors) can indicate 'higher' or 'lower' biological age, which has been linked to age-related health outcomes, including mortality.

🥇 1st Place: Julian Reinhard, also known as “DarthVenter,” Machine Learning Scientist at Evotec, achieved a final score of 2.45 years age error. Julian received $15,000 USD in cash prizes.
🥈 2nd Place: Lucas Paulo de Lima Camillo, Head of Machine Learning at Shift Bioscience, achieved a final score of 2.55 years age error. Lucas received $10,000 USD in cash prizes.
🥉 3rd Place: Team “ZetaPartition”, comprising academics Jakob Träuble and Stefan Jokiel, achieved a final score of 2.46 years age error. The team received $5,000 USD in cash prizes.

Phase II - Mortality Age

Phase II leveraged the same unique, high quality dataset as in the first phase, and pushed the boundaries of aging research by asking participants to develop cutting-edge proteomic models predicting mortality risk.

🥇 1st Place: "Team Kauai"
Comprised of Jakob Träuble of University of Cambridge, Korbinian Träuble of Helmholtz Munich, and Raphael Lermer of Deutsches Herzzentrum München, Team Kauai developed a proteomic model that outperformed the GrimAgeV2 epigenetic clock. Their work sets a new standard for innovation in aging biomarker research and their model provides significant advancements in predicting mortality and age-related diseases.

Join Phase 3 (Multi-Morbidity) of the Challenge!