MBZUAI Human Phenotype Project: Landmark Mapping of Health Evolution and Disease Onset

Unlocking the Health-Disease Continuum Through Deep Phenotyping

The Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) has positioned itself at the forefront of precision medicine with its groundbreaking Human Phenotype Project (HPP). This ambitious initiative delves into the intricate dynamics of human health, mapping how biological, environmental, and lifestyle factors interplay to influence disease onset and progression. By leveraging artificial intelligence (AI) on vast multimodal datasets, HPP is redefining preventive healthcare, offering insights that could transform how we approach chronic conditions like diabetes and cardiovascular disease (CVD).

Launched in 2022, HPP represents a paradigm shift from reactive treatment to proactive intervention. Researchers at MBZUAI, in collaboration with global partners like the Weizmann Institute of Science, have enrolled nearly 28,000 participants worldwide, with over 13,000 completing in-depth profiling. This longitudinal effort, spanning up to 25 years, aims to create 'digital twins'—personalized AI simulations of an individual's health trajectory—for tailored interventions.

Origins and Vision of the Human Phenotype Project

MBZUAI's HPP emerged from a vision to bridge the gap between genomics and real-world health outcomes. Led by Professor Eran Segal, now Acting Dean of the School of Digital Public Health at MBZUAI, the project addresses the limitations of static snapshots like single blood tests. Instead, it captures the 'phenotype'—observable traits shaped by genes, environment, and behavior—through continuous, high-resolution data.

The UAE's investment in AI-driven research underscores MBZUAI's role as a hub for innovation. With Abu Dhabi's focus on diversifying beyond oil into knowledge economies, projects like HPP align with national goals for advanced healthcare. The university's computational biology expertise enables the integration of diverse data streams, fostering discoveries applicable to the region's high prevalence of metabolic diseases.

Initial targets of 10,000 participants have expanded globally, powered by Pheno.AI for standardized data collection. This scale allows for ethnically diverse cohorts, crucial for establishing norms that reflect real population variations.

The Multimodal Data Revolution

HPP's strength lies in its unprecedented data depth: over 30 modalities per participant. These include:

• Medical history and family genetics.
• Lifestyle, nutrition (over 3 million meals logged), and physical activity from wearables.
• Vital signs, anthropometrics, and continuous glucose monitoring (CGM) yielding 10 million+ readings.
• Sleep tracking, imaging (fundus, carotid ultrasound), and multi-omics: genetics, transcriptomics, microbiomes (gut, oral, vaginal), metabolomics, immune profiling.

This holistic approach reveals hidden patterns. For instance, CGM data exposed day-to-day fasting glucose variability, reclassifying 40% of 'normal' individuals as prediabetic—a finding that challenges traditional diagnostics.

Privacy is paramount: all data is de-identified, with AI models trained via self-supervised learning to predict outcomes without compromising anonymity.

Landmark Nature Medicine Findings: Mapping Health Evolution

Published July 15, 2025, in Nature Medicine, the HPP study (DOI: 10.1038/s41591-025-03790-9) unveils the health-disease continuum. Co-senior authored by MBZUAI President Eric Xing and Eran Segal, it analyzes 13,000+ profiled participants.

Key revelation: phenotypes vary markedly by age and ethnicity, urging personalized norms for blood tests and behaviors. Biological aging—measured via cardiovascular metrics and RNA-seq—outpredicts chronological age for risks like elevated triglycerides and HbA1c.

Ultra-processed foods (UPF) correlate with higher BMI, blood pressure, and reduced microbiome diversity, while Mediterranean or vegan diets yield favorable metrics. Microbiome signatures distinguish breast cancer, IBD, and endometriosis patients from controls.

Spotlight on Chronic Diseases: Obesity, Diabetes, and CVD

HPP pinpoints drivers of metabolic ills. Obesity links to UPF-driven visceral fat, accelerating aging. Diabetes insights from CGM show overlooked variability, with AI flagging prediabetes early.

For CVD, the COMPRER model fuses imaging for superior risk prediction. Metabolomics in pancreatic cancer reveals individualized mechanisms, advocating subtype-specific therapies.

"By capturing how health evolves day by day, HPP enables actionable precision medicine," notes Segal.

GluFormer: AI Powerhouse from HPP Data

A companion Nature paper (Jan 2026) introduces GluFormer, trained on HPP's CGM data. This transformer-based foundation model outperforms HbA1c and GMI, forecasting diabetes 12 years ahead (66% new cases in high-risk group) and CVD death (69% events captured).

Integrating diet data enhances meal-time accuracy, simulating personalized responses. Developed with NVIDIA and Weizmann, it exemplifies HPP's AI translation.

Explore careers in AI-health at higher-ed-jobs or UAE opportunities via AcademicJobs UAE.

AI-Driven Digital Twins and Precision Medicine

HPP's multimodal AI fuses data into digital twins, simulating interventions like diet changes. This 'personalized blueprint' predicts trajectories, shifting UAE healthcare toward prevention amid rising diabetes rates.

Xing emphasizes: "HPP is the dawn of true precision medicine."

For researchers, HPP data accelerates biomarker discovery; students can engage via MBZUAI programs. Check academic CV tips.

UAE's Leadership: MBZUAI and National Impact

In the UAE, where metabolic diseases burden healthcare, HPP aligns with Vision 2031. MBZUAI's Institute of Digital Public Health analyzes UAE genotypes, informing local strategies.

Collaborations with Weizmann and NVIDIA amplify impact, positioning Abu Dhabi as an AI-health nexus. Internal links: Abu Dhabi university jobs.

Challenges, Ethics, and Future Horizons

Challenges include scaling diverse cohorts and ethical AI use. HPP prioritizes privacy, de-identifying data for global sharing.

Future: Full digital twins, clinical trials in 1-3 years, practice-wide adoption in 5-10. HPP could cut UAE disease burdens, inspiring research jobs.

Photo by Irfannur Diah on Unsplash

Career Opportunities in AI-Health Research

HPP exemplifies UAE higher ed's edge. Aspiring AI biologists? MBZUAI offers programs blending computation and biomedicine. Visit Rate My Professor, higher-ed-jobs, university-jobs, higher-ed-career-advice, and post-a-job for openings. Engage via comments below.