Driving Discovery in Spatial Proteomics: Protein Sequencing, Domain-Adaptive Imaging, and Generative AI
Jacob Luber, UT-Arlington
Time: 9:00am Monday March 17 2025 {San Francisco, CA, USA}
Abstract:
Proteins are the molecular engines and master regulators of biological function. However,
current technologies for protein sequencing and immunohistopathological imaging still face
significant limitations in comprehensiveness, throughput, and cost. In this talk, I will discuss
my lab’s two overarching research thrusts. First, we are developing a next-generation protein
sequencing platform by integrating click-chemistry–based experimental protocols with large
language models, allowing partial amino acid readouts to be reconstructed into full-length
protein sequences [1]. Second, we are addressing the challenge of domain shift in computational
pathology by building robust, machine-learning–driven pipelines for histopathology
slide indexing and retrieval under diverse imaging conditions [2]. I will also highlight our
work at the intersection of these two areas, where generative AI methods enable synthesis
of realistic multiplexed biomarker channels in spatial proteomics, augmenting experimental
datasets and advancing our understanding of tissue microenvironments [3]. By bridging
state-of-the-art experimental techniques with modern deep learning strategies, we seek to
push the boundaries of spatially resolved protein biology for applications in precision diagnostics
and targeted therapeutics.
Jacob Luber:
Jacob M. Luber, Ph.D. is an Assistant Professor of Computer Science and Engineering at the
University of Texas at Arlington (UTA) and an Affiliate Assistant Professor of Bioengineering.
He directs a multidisciplinary research group that integrates bioinformatics, machine
learning, and cancer genomics to address challenges in immunology, tumor heterogeneity,
and clinical diagnostic workflows.
1
Dr. Luber received his Ph.D. in Biomedical Informatics from Harvard University, where
his dissertation focused on systems-level interrogation of host–microbiome interactions in
disease. He completed a postdoctoral fellowship at the National Cancer Institute, National
Institutes of Health, working in the Cancer Data Science Laboratory. Since joining UTA, he
has served as a faculty affiliate at the Multi-Interprofessional Center for Health Informatics
and has received multiple competitive grants, including a Cancer Prevention and Research
Institute of Texas (CPRIT) award and a University of Texas System Rising STARs Award,
together totaling over $2.5 million as lead investigator. His academic contributions have
garnered over 4000 citations, reflecting his leadership in computational biology, spatial proteomics,
and applied artificial intelligence.
In his current role, Dr. Luber teaches graduate and undergraduate courses in bioinformatics
and special topics in AI-driven medical imaging. Alongside his academic efforts,
he maintains active collaborations with clinical scientists and industry partners, ensuring
that cutting-edge research in protein sequencing, histopathology imaging, and generative AI
rapidly translates to practical impacts on patient care.
Selected References
[1] Pham TLH, Saurav JR, Omere AA, et al. Peptide Sequencing Via Protein Language
Models. In: Proceedings of the 15th ACM Conference on Bioinformatics, Computational
Biology, and Health Informatics (ACM-BCB). 2024.
[2] Shang HH, Nasr MS, Veerla JP, et al. Histopathology Slide Indexing and Search—Are
We There Yet? NEJM AI. 2024;1(5).
[3] Saurav JR, Nasr MS, Shang HH, et al. A SSIM Guided cGAN Architecture For Clinically
Driven Generative Image Synthesis of Multiplexed Spatial Proteomics Channels. In: 2023
IEEE Conference on Computational Intelligence in Bioinformatics and Computational
Biology (CIBCB). IEEE, 2023.
|
|
