UTA researchers secure $1.96 million to develop AI tools for eye-disease research

Researchers studying eye disease could gain a way to identify promising biological targets without running expensive single-cell experiments for every regulator. The University of Texas at Arlington announced October 8 that its researchers and UT Southwestern collaborators are developing two AI-powered tools through a four-year project funded by a $1.96 million National Eye Institute grant.

The project focuses on biological regulators—the “control switches” UTA describes as helping keep eyes healthy and contributing to disease. Its ambition is to give scientists a clearer view of those regulators across whole tissues and within particular cell types, helping them decide which targets deserve further investigation.

Xinlei (Sherry) Wang, a professor of statistics and data science in UTA’s mathematics department, is collaborating with Lin Xu, an assistant professor at UT Southwestern. Their teams bring together statistical and computational expertise with ocular biology and clinical experience.

Finding regulators in large datasets

The first planned tool will analyze large genomic datasets to identify regulators important to eye health and disease. Wang says the algorithm must be scalable: it needs to run smoothly as the collections of data it examines become very large.

That work builds on Wang’s research into using AI and advanced statistical methods to uncover biological patterns hidden in massive datasets. The intended benefit is a more focused search: Wang says pinpointing key regulators would help researchers concentrate on the most promising targets, saving time and resources.

When researchers analyze a whole tissue sample, the signals are mixed together.

Xinlei (Sherry) Wang, UTA professor of statistics and data science

Untangling signals without testing every cell separately

The second tool addresses a different obstacle. Eye tissue contains many cell types with distinct functions, so analyzing a whole sample combines their signals. The team plans to develop a deconvolution tool—a computational method for separating those mixed signals—to examine how a regulator behaves in specific cell types.

Wang says that approach would avoid requiring very expensive single-cell experiments for every regulator. The proposed shortcut concerns how researchers study regulators; the two tools are still being developed, and their stated purpose is to support research into disease mechanisms.

The long-term goal is to improve understanding of eye disease and identify promising directions for future diagnostics and targeted therapies. Those applications are the destination of the research, not treatments or diagnostic products being released with the grant announcement.