Using Empire AI to Uncover Genetic Variations' Influence over Health

Using Empire AI to Uncover How Genetic Variations Influence Health and Disease and Accelerate Discovery of Disease Treatments

Researcher:  Li Shen, Professor of Neuroscience
Institution:  The Icahn School of Medicine at Mount Sinai

The Challenge:

Every cell in the human body contains a copy of the genome, yet scientists still do not fully understand how a large portion of that genetic code works. To address this issue, researchers from the Icahn School of Medicine at Mount Sinai are developing OmniGenome, a powerful AI system designed to help decode the genome and uncover how genetic variation influences health and disease.

The challenge is that the genome contains billions of genetic letters, and many of the DNA changes associated with disease occur in regions whose functions remain poorly understood.

By learning from vast amounts of genomic data, OmniGenome aims to help scientists identify which genetic changes influence disease, uncover new biological mechanisms, and accelerate discoveries across the biomedical research community. Ultimately, researchers hope these advances will speed the path from scientific discovery to new diagnostics, treatments, and cures.

The approach: 

Just as large language models have transformed how we understand and generate text, genome AI models have the potential to transform how we understand biology. These models can learn patterns from massive amounts of genomic data and help scientists predict how DNA influences cellular behavior, disease risk, and treatment response.

OmniGenome is designed to analyze extremely long stretches of DNA and learn patterns from large genomic datasets. By connecting DNA sequences with biological activity observed across large datasets of tissues, cell types, and disease states, the model can identify relationships that would be difficult to discover using traditional research methods.  The model is designed to predict how genes are turned on and off in different tissues and cell types, helping researchers understand how genetic variation can lead to disease.

Why Empire AI Matters

Training state-of-the-art genome AI models requires enormous computational resources. Empire AI enables us to train larger models, process more data, and perform experiments at a scale that would otherwise be out of reach for many academic labs.

Most existing genomic AI models focus on a limited set of biological measurements or specific prediction tasks. This work aims to build more general and comprehensive models that can learn many aspects of genome function simultaneously.  It is also developing open and transparent models that can be broadly used, scrutinized, and improved by the scientific community.

By providing world-class computing resources to academic researchers, Empire AI helps to ensure that breakthroughs in genome AI are developed in an environment committed to open science, collaboration, and public benefit.

What Researchers are Learning

If successful, genome AI could help researchers identify disease-causing genetic variants, better understand complex diseases such as cancer and neurological disorders, and accelerate the development of new therapies. More broadly, it could provide a powerful new framework for studying human biology and important tools to help scientists generate hypotheses and accelerate research.

Researchers believe biology is entering a transformation similar to the one AI has already brought to language and computer vision. By combining large-scale biological data with advanced AI, projects like OmniGenome could help accelerate discoveries in genetics, disease mechanisms, drug development, and precision medicine for years to come.

Impact for New York

Advances in genome AI can strengthen New York’s leadership in biotechnology and life sciences by supporting cutting-edge research, attracting investment, and accelerating the development of new diagnostics and therapies. The project also helps to train the next generation of AI and biomedical researchers within New York’s academic institutions.