Seminars

Abstract: Resistance to cancer therapy arises from tumors’ remarkable ability to adapt to diverse and challenging conditions. This adaptation occurs not only through accelerated Darwinian evolution but also by leveraging the inherent plasticity of multicellular tissues. At the core of this plasticity is the dynamic cellular composition of tumors, shaped by multiple concurrent biological processes and pressures. Dissecting this complexity and linking it to clinical outcomes requires experimental and computational approaches capable of resolving cell states with high resolution and across large cohorts of patients.

In this talk, I will share our ongoing work combining single-cell and bulk omics data with novel computational deconvolution approaches to reveal the role of the tumor microenvironment in driving the mesenchymal transformation of ependymoma and other brain cancers. Central to our efforts is ConDecon, a computational method that infers the likelihood of each cell from a reference single-cell dataset being present in a bulk tissue sample, without relying on predefined labels or cell-type–specific gene expression signatures. By applying ConDecon and related tools, we aim to better characterize the cellular landscape underlying tumor plasticity, ultimately providing new insights into therapeutic resistance and potential strategies for intervention.

Bio: Pablo G. Camara is an Associate Professor of Genetics at the University of Pennsylvania and a faculty member of the Institute for Biomedical Informatics and the AI2D Center for AI and Data Science for Integrated Diagnostics. His research explores the cellular and molecular organization of the brain and brain tumors, using mathematical principles to analyze high-dimensional omics and imaging data. Dr. Camara received a Ph.D. in Theoretical Physics from the Universidad Autonoma de Madrid in 2006 and continued his postdoctoral work at Ecole Polytechnique (France), the European Organization for Nuclear Research (CERN, Switzerland), and the University of Barcelona. Fascinated by fundamental questions in biomedicine, he shifted his focus to quantitative biology in 2014 as a postdoctoral fellow at Columbia University's Systems Biology Department and a research associate at the Institute for Advanced Study in Princeton. Since joining the University of Pennsylvania as a faculty member in 2018, his group has developed geometry- and topology-based algorithms for integrating and analyzing single-cell multi-omics, cytometry, and imaging data, and is using them to elucidate the cellular ecosystem and oncogenic pathways of glioma and to characterize CAR-T cell immunotherapies functionally.

Abstract: Medical artificial intelligence has shown remarkable progress in diagnostic and clinical applications, yet concern grows over AI systems relying on "shortcut learning"—exploiting statistical patterns that correlate with outcomes without capturing genuine causal relationships. This paper examines the dual nature of shortcut learning in medical AI, analyzing both its potential benefits and significant risks. While shortcuts can improve computational efficiency and enable rapid deployment in resource-constrained settings, they simultaneously present substantial threats to generalization, fairness, and reliability when deployed in heterogeneous clinical environments. We review recent examples of shortcut learning in medical imaging, electronic health record analysis, and clinical decision support systems, presenting a framework for distinguishing between benign shortcuts that enhance performance and harmful shortcuts that undermine clinical utility. This analysis is complemented by proposed methodological solutions for detecting and mitigating problematic shortcuts, including causality-informed architectures, diverse training datasets, and human-AI collaborative validation. We conclude that shortcut learning represents both opportunity and threat—with the balance determined by how effectively the medical AI community addresses this fundamental challenge through rigorous development practices, transparent evaluation, and ongoing clinical oversight.

Bio:

Dr. Gichoya is an associate professor at Emory university in Interventional Radiology and Informatics leading the HITI (Healthcare AI Innovation and Translational Informatics) lab . Her work is centered around using data science to study health equity. Her group works in 4 areas - building diverse datasets for machine learning (for example the Emory Breast dataset); evaluating AI for bias and fairness; validating AI in the real world setting and training the next generation of data scientists (both clinical and technical students) through hive learning and village mentoring. She serves as the program director for radiology:AI trainee editorial board and the medical students machine learning elective. She has mentored over 60 students across the world (now successful faculty, post doc, PHD and industry employees) from several institutions around the world. She has received several awards including the most influential radiology researcher in 2022, and is a 2023 Emerging Scholar in the National Academy of Medicine. 

Abstract: Physical health and chronic medical comorbidities are underestimated, inadequately treated, and often overlooked in psychiatry. Integrated research into brain and body systems holds substantial clinical potential in addressing mental-physical comorbidity. In my talk, I will first present evidence of co-occurrent physical health problems in people with mental disorders, from epidemiology to physiology. I will then present several lines of evidence exploring the underlying mechanisms explaining the development and progress of mental-physical comorbidity, including aging, immunometabolic dysfunction and lifestyle changes and the key role of the brain in the intertwined relationships. To close the talk, I will discuss the way forward to reduce the adverse effect of physical comorbidity in people with mental illness.

Bio:

Dr Ye Ella Tian is a National Health and Medical Research Council (NHMRC)-funded Senior Research Fellow at the Department of Psychiatry, The University of Melbourne, Australia. She is a psychiatrist and neuroscientist by training and holds a PhD in systems neuroscience. During her PhD, she develops the now widely used Melbourne Subcortex Atlas.

She currently holds an NHMRC Investigator Grant, investigating brain-body relationships in mental illness across the lifespan. She works at the interface between neuroscience, computation and translational research of applying brain imaging techniques to clinical research.

Abstract: Reliable and accurate prediction tools are essential for advancing precision medicine. With the rapid accumulation of diverse biomedical data—both in scale and variety—there is a growing opportunity to leverage these data for clinical decision making. This talk highlights advancements in breast cancer prognosis and polygenic risk prediction through large-scale data integration. First, we validated the PREDICT Breast v3 prognostic model in a cohort of over 860,000 breast cancer patients, demonstrating robust overall performance while identifying areas for model refinement. Building on these insights, we develop PREDICT Breast v4, a machine learning-based model that integrates clinical and socioeconomic data to improve predictive accuracy. Additionally, we introduce the S4-Multi model, a cross-ancestry polygenic risk model designed for multiple phenotypes and biobanks. Together, these studies highlight the potential of integrating diverse datasets to improve risk assessment, inform clinical decisions, and optimize personalized treatment strategies.

Bio: Dr. Pei-Chen Peng is an Assistant Professor in the Department of Computational Biomedicine at Cedars-Sinai Medical Center. Her research focuses on machine learning and statistical modeling of heterogenous multi-omics data to improve the prevention and treatment of cancer and other diseases. She obtained her Ph.D. in Computer Science from University of Illinois at Urbana-Champaign and holds a M.S. and a B.S. in Computer Science from National Taiwan University. She received the NIH/NCI Early K99/R00 Pathway to Independence Award in 2021 and was recognized as a Rising Star in Electrical Engineering and Computer Sciences. 

Abstract: TBA