Upcoming Event: Oden Institute Seminar

Phenotype-Guided AI for Large-Scale Single-Cell Analysis of the Tumor Microenvironment

Zheng Xia, Associate Professor, The University of Texas at Austin

Abstract

Single-cell sequencing (scSeq) enables the identification of diverse cell subpopulations within heterogeneous tissue ecosystems, providing insights into biological and clinical mechanisms. However, interpreting complex single-cell data remains challenging. Identifying key subpopulations driving clinical phenotypes, such as tumor metastasis, treatment resistance, and survival outcomes, is crucial for advancing targeted therapies and biomarker discoveries. To address this challenge, we first developed a novel framework called SCISSOR, which integrates widely available bulk phenotype information with single-cell data to identify disease-relevant subpopulations for cellular target discovery. Secondly, in single-cell experiments designed to profile samples from diverse conditions (e.g., treatment-resistant vs. responder groups), identifying subpopulations unique to each phenotype is essential for enhancing phenotype-specific gene signal detection. To this end, we developed a clustering-free method using a novel learning-with-rejection platform to detect high-confidence, phenotype-enriched subpopulations from millions of cells. Finally, we propose using attention-based multiple instance learning to perform single cell survival analysis, a first of its kind.

Biography

Zheng Xia is an Associate Professor in the Department of Biomedical Engineering at The University of Texas at Austin and affiliated faculty in the Oden Institute for Computational Engineering and Sciences. Xia’s lab develops machine learning and artificial intelligence methods to transform large-scale biomedical data into biological insights and clinically actionable predictions, with a focus on linking molecular measurements to clinical phenotypes using single-cell, spatial, and multimodal data. His group has developed bioinformatics methods published in Nature Biotechnology, Nature Machine Intelligence, and Cancer Discovery. Xia joined UT Austin as a CPRIT Scholar in Cancer Research, and his research is supported by the NIH and the Department of Defense.