AI Tool PhyloFrame Tackles Ancestral Bias in Genetic Research for Improved Precision Medicine

University of Florida Researchers Develop AI Tool to Address Ancestral Bias in Genetic Research

Researchers at the University of Florida have developed a groundbreaking AI tool called PhyloFrame to tackle a critical issue in medical genetic research: ancestral bias. Led by Dr. Kiley Graim, an assistant professor in the Department of Computer & Information Science & Engineering, the team aims to improve precision medicine outcomes for people of all backgrounds 1

1

.

The Problem of Ancestral Bias

Ancestral bias in genetic data occurs when most research is based on data from a single ancestral group, typically of European descent. Dr. Graim estimates that 97% of sequenced samples come from people of European ancestry, due to funding priorities and socioeconomic factors 2

2

. This bias limits advancements in precision medicine and leaves large portions of the global population underserved in disease treatment and prevention.

PhyloFrame: An AI Solution

PhyloFrame is a machine-learning tool that uses artificial intelligence to account for ancestral diversity in genetic data. The tool integrates massive databases of healthy human genomes from the population genomics database gnomAD with smaller disease-specific datasets used to train precision medicine models 3

3

.

Key features of PhyloFrame include:

1. Ability to predict differences between disease subtypes
2. Suggestion of best treatments for patients regardless of ancestry
3. Improved accuracy across diverse populations

The Development Process

The development of PhyloFrame involved processing enormous amounts of genetic data. The team utilized UF's HiPerGator, one of the most powerful supercomputers in the country, to analyze genomic information from millions of people, processing 3 billion base pairs of DNA for each individual 1

1

.

Implications for Precision Medicine

PhyloFrame's ability to consider genetic differences linked to ancestry sets it apart from current models. This is crucial because existing data often comes from research hospitals and patients who trust the healthcare system, leaving out populations in small towns or those who distrust medical systems 2

2

.

Dr. Graim believes that tools like PhyloFrame will eventually be used in clinical settings, replacing traditional models to develop treatment plans tailored to individuals based on their genetic makeup 3

3

.

Future Directions

The team's next steps include:

1. Refining PhyloFrame
2. Expanding its applications to more diseases
3. Developing more sophisticated models
4. Refining how populations are defined

The project received funding from the UF College of Medicine Office of Research's AI2 Datathon grant award, supporting the use of AI tools to improve human health 2

2

.

As countries like China and Japan work to close the data gap, PhyloFrame represents a significant step towards more equitable and effective precision medicine, potentially revolutionizing how diseases are predicted, diagnosed, and treated for diverse populations worldwide.