Spatial Artificial Intelligence

Computational Medicine

Zheng Lab

Spatial Artificial Intelligence

We develop next-generation multimodal AI for modeling spatially organized biological systems, integrating imaging, omics, and sensing data to advance precision medicine.

Explore research

Mission

Our mission is to develop computational methods to understand how cells and tissues are organized, interact, and change in health and disease. By integrating multimodal biomedical data, we aim to uncover mechanisms of disease initiation, progression, and therapeutic response, and translate these insights into better patient stratification and treatment selection.

Mission illustration showing multimodal AI across biological and physical scales
Yuanning Eric Zheng headshot

Principal Investigator

Yuanning (Eric) Zheng, PhD

Dr. Yuanning Zheng is a computational scientist who develops quantitative methods to understand tissue biology in health and disease. His research integrates medical imaging, genomics, single-cell profiling, and spatial omics to characterize tissue ecosystems and uncover mechanisms of disease progression and therapeutic response.

Dr. Zheng received his M.S. in Computer Science from the Georgia Institute of Technology and his Ph.D. in Medical Sciences from Texas A&M University. He completed his postdoctoral training at Stanford University, where he was mentored by Olivier Gevaert and Zinaida Good.

His research has received support from the National Cancer Institute through the NIH K99/R00 Pathway to Independence Award. Dr. Zheng maintains active collaborations with clinicians, pathologists, engineers, and AI scientists across academia and medicine, with the goal of translating computational innovations into real-world impact in precision medicine.

Contact: [email protected] LinkedIn Google Scholar ORCID

Research

Core directions

01

Spatial Intelligence for Biological Systems

Cells are the fundamental units of life. Within tissues, however, cells do not function in isolation. Instead, they form complex physical and signaling interactions with neighboring cells and their surrounding microenvironment. Dysregulation of these interaction patterns can drive disease initiation, progression, and resistance to therapy.

We develop computational methods to quantitatively characterize cellular interactions, tissue architecture, and spatial ecosystems in health and disease. By integrating single-cell and spatial profiling technologies, including single-cell RNA sequencing, CODEX, CosMx, and Visium, we seek to understand the functional states and spatial organization of cells within tissue microenvironments. We have a particular interest in antigen-presenting cells and T cells, and in determining how their cellular states and interactions influence disease progression and therapeutic response. A major area of application is T-cell-dependent cancer immunotherapy, including CAR-T therapy and immune checkpoint blockade.

Illustration of spatial AI for molecular and biological systems
02

Foundation Models for Medical Imaging

Radiology and histopathology images are widely available diagnostic assays routinely collected from cancer patients. These images contain rich information about tissue morphology, texture, cellular composition, and topological organization. However, this information is frequently underutilized because disease-related patterns are often subtle and complex, making them difficult to quantify consistently through human observation alone.

We develop representation learning strategies to extract, quantify, and interpret features from medical images and integrate them with genomic, transcriptomic, and proteomic information. Our goal is to enable mechanistic discovery, improve diagnosis, and guide treatment allocation.

Illustration of foundation models for multimodal medical imaging
03

Multimodal AI for Precision Biomedicine

Human diseases emerge from complex interactions across molecular, cellular, tissue, organ, and clinical scales. No single data modality can fully capture this complexity. Genomic, transcriptomic, spatial, imaging, clinical, and language data each provide complementary views of disease biology and patient trajectories.

We develop multimodal AI frameworks to integrate diverse biomedical data types, including omics profiles, spatial measurements, medical images, electronic health records, and clinical text. By connecting biological mechanisms with patient-level outcomes, we aim to build predictive and interpretable models that improve disease subtyping, treatment allocation, and therapeutic monitoring.

Illustration of multimodal AI for precision biomedicine

Software

Platforms

NucSegAI

A deep-learning model that leverages H&E-stained histology images to predict response to anti–PD-1 immunotherapy in non-small cell lung cancer by characterizing cellular morphology and spatial tissue organization.

paper | website

SEQUOIA

AI-powered digital pathology platform that predicts transcriptomic profiles from whole-slide images of cancer tissues at bulk and locoregional levels.

paper | website

EpiMix

R-based computational tool for population-scale analysis of epigenomic and transcriptomic data across promoters, enhancers, microRNAs, lncRNAs, and genes.

paper | website

Selected publications

Recent work

First page of Science Advances 2025 publication

Zheng Y, Sadée C, Ozawa M et al. Single-cell multi-modal analysis reveals tumor microenvironment predictive of treatment response in non-small cell lung cancer. Science Advances, 11(21), 2025.

First page of Nature Communications 2024 SEQUOIA publication

Pizurica M*, Zheng Y*, Carrillo-Perez F et al. Digital profiling of cancer transcriptomes from histology images with linearized vision attention. Nature Communications, 15(9886), 2024.

First page of Cell Reports Methods 2023 EpiMix publication

Zheng Y, Jun J, Brennan K et al. EpiMix is an integrative tool for epigenomic subtyping using DNA methylation. Cell Reports Methods, 3, 100515, 2023.

First page of Nature Communications 2023 GBM360 publication

Zheng Y, Carrillo-Perez F, Pizurica M et al. Spatial cellular architecture predicts prognosis in glioblastoma. Nature Communications, 14(4122), 2023.

First page of Nature Medicine 2023 publication

Thieme AH, Zheng Y, Machiraju G et al. A deep-learning algorithm to classify skin lesions from mpox virus infection. Nature Medicine, 29, 738-747, 2023.

First page of Nature Biomedical Engineering 2024 publication

Carrillo-Perez F, Pizurica M, Zheng Y et al. Generation of synthetic whole-slide image tiles of tumours from RNA-sequencing data via cascaded diffusion models. Nature Biomedical Engineering, 2024.

First page of PLOS Computational Biology Moonlight publication

Nourbakhsh M, Zheng Y, Noor H et al. Revealing cancer driver genes through integrative transcriptomic and epigenomic analyses with Moonlight. PLOS Computational Biology, 21(4), e1012999, 2025.

First page of PLOS Computational Biology reliability publication

Zhan X, Xu Q, Zheng Y et al. Reliability-based cleaning of noisy training labels with inductive conformal prediction in multi-modal biomedical data mining. PLOS Computational Biology, 21(2), e1012803, 2025.

First page of npj Breast Cancer 2025 publication

Noor H, Zheng Y, Mantz A et al. A 20-feature radiomic signature of triple-negative breast cancer identifies patients at high risk of death. npj Breast Cancer, 11, 79, 2025.

First page of Communications Medicine 2023 brain tumors publication

Steyaert S, Qiu YL, Zheng Y et al. Multimodal deep learning to predict prognosis in adult and pediatric brain tumors. Communications Medicine, 3, 44, 2023.

First page of Frontiers in Oncology 2024 publication

Brooks J*, Zheng Y*, Hunter K et al. Digital Spatial Profiling identifies distinct patterns of immuno-oncology-related gene expression within oropharyngeal tumours in relation to HPV and p16 status. Frontiers in Oncology, 14, 2024.

Join Us

Open positions

We are recruiting highly motivated PhD students, research technicians, and postdoctoral fellows to join our interdisciplinary research program at the intersection of computation, biology, and medicine. Our lab develops computational approaches to study tissue organization and disease biology using spatial omics, single-cell genomics, medical imaging, and other multimodal biomedical data. Successful candidates will have:

  1. Strong programming and quantitative skills in languages such as Python or R;
  2. Experience with machine learning, deep learning, computational biology, bioinformatics, image analysis, or related computational methods;
  3. Experience working with biomedical imaging, genomic, spatial omics, or other large-scale biological datasets is preferred;
  4. A strong interest in applying computational methods to biologically and clinically relevant problems;
  5. Intellectual curiosity, a collaborative mindset, and strong communication skills.

Candidates from backgrounds including computer science, machine learning, computational biology, bioinformatics, biomedical engineering, medical imaging, data science, biostatistics, and related quantitative disciplines are encouraged to apply.

Interested applicants should send the following materials to the PI's email address at [email protected]:

  1. Curriculum vitae (CV);
  2. A cover letter describing their research experience, technical background, and long-term career goals;
  3. Contact information for three references, including their relationship to the applicant.

PhD student

Qualification: Recently completed or currently completing a bachelor's or master's degree in computer science, biomedical engineering, bioinformatics, data science, biostatistics, biomedical sciences, or a related field.

Research technician

Qualification: Bachelor's degree in a quantitative, computational, or biomedical discipline, with programming experience in Python or R. Prior experience in data analysis, machine learning, biomedical imaging, genomics, or computational biology is preferred but not required.

Postdoctoral scholar

Qualification: PhD in a quantitative, computational, or biomedical discipline, completed within the past five years or expected before the start date. A strong research record, including at least one first-author publication, is preferred.

Why join us

  1. Structured mentoring and accessible support for research, manuscript preparation, and presentations;
  2. Support for attending conferences and building professional networks;
  3. Support for applying to scholarships and fellowships, including F and K awards from the National Institutes of Health (NIH) and other funding resources;
  4. A strong commitment to work-life balance.

Alumni

Current and previous trainees

  • Jane Feng (Stanford Cancer Immunology Program)
  • Clove Tayler (Stanford Cancer Immunology Program)
  • Alexa Chen (Stanford Institutes of Medicine Summer Research Program)
  • Keshav Narang (Stanford Institutes of Medicine Summer Research Program)
  • Arnoldo Sanchez (Stanford-Foothill STEM Internship Program)
  • Jessie Altamirano (Stanford-Foothill STEM Internship Program)
  • John Jun (Undergraduate research program)
  • Markus Sujansky (Undergraduate research program)

Contact

Collaborations

We welcome collaborations and discussions with clinicians, pathologists, biologists, and industrial partners who are interested in applying spatial AI and multimodal data to advance precision medicine and healthcare.

Yuanning (Eric) Zheng
[email protected]
LinkedIn

Institutional Home

University of Houston

University of Houston

The University of Houston is a leading public research university in Houston, one of the largest universities in Texas, and a Carnegie-designated Tier One public research institution. Located in a city shaped by medicine, energy, engineering, and global industry, UH provides a strong environment for translational research that connects computational innovation with real-world impact.

About UH

Department of Electrical and Computer Engineering

The Department of Electrical and Computer Engineering at the Cullen College of Engineering advances research and education across imaging, communication technologies, computer and information systems, power and energy systems, control systems, and electromagnetics. This multidisciplinary setting aligns naturally with the lab's work in spatial intelligence, multimodal AI, medical imaging, and complex biological and physical systems.

UH ECE