Artifical Intelligence in Translational Systems Medicine (Ergen Lab)
Research
Patients with the same clinical diagnosis experience markedly different disease courses and respond differently to identical therapies. These differences arise from complex molecular processes spanning individual cells, tissues, and the organism as a whole, yet current clinical classifications capture only a fraction of this biological diversity. Our laboratory investigates the molecular basis of disease heterogeneity by developing artificial intelligence methods that integrate information across biological scales. Our long-term goal is to identify the molecular mechanisms that define clinically relevant disease subtypes and translate these insights into more precise diagnosis and treatment.
Recent advances in single-cell genomics, spatial omics, and digital health have transformed our ability to characterize human disease. Rather than studying individual measurements in isolation, these technologies now make it possible to record comprehensive molecular representations of patients including cellular composition, tissue architecture, and clinical information. At the same time, modern generative AI and machine learning provide new opportunities to integrate these highly heterogeneous data into predictive models of disease biology. Together, these developments are laying the foundation for a new generation of data-driven systems medicine.
Our laboratory develops computational methods that combine multimodal molecular profiling, digital pathology, and longitudinal clinical data to study human disease across biological scales. We investigate how molecular states emerge, evolve during disease progression, and determine therapeutic response. To understand the biological mechanisms underlying these computational predictions, we integrate AI-based analyses with experimental validation in complex in vitro systems. Through this interdisciplinary approach, we aim to establish a mechanistic and predictive framework for precision medicine that enables more accurate diagnosis, patient stratification, and individualized treatment.
ResolVI - addressing noise and bias in spatial transcriptomics, Can Ergen, Nir Yosef
bioRxiv 2025.01.20.634005; doi: https://doi.org/10.1101/2025.01.20.634005
Boyeau, P., Hong, J., Gayoso, A. et al. Deep generative modeling of sample-level heterogeneity in single-cell genomics. Nat Methods 22, 2264–2274 (2025).https://doi.org/10.1038/s41592-025-02808-x
Wells, S.B., Rainbow, D.B., Mark, M. et al. Multimodal profiling reveals tissue-directed signatures of human immune cells altered with age. Nat Immunol 26, 1612–1625 (2025). https://doi.org/10.1038/s41590-025-02241-4
Ergen, C., Pour Amiri, V.V., Kim, M. et al. Scvi-hub: an actionable repository for model-driven single-cell analysis. Nat Methods 22, 1836–1845 (2025). https://doi.org/10.1038/s41592-025-02799-9
Ergen, C., Xing, G., Xu, C. et al. Consensus prediction of cell type labels in single-cell data with popV. Nat Genet 56, 2731–2738 (2024). https://doi.org/10.1038/s41588-024-01993-3
Tabula Sapiens reveals transcription factor expression, senescence effects, and sex-specific features in cell types from 28 human organs and tissues, The Tabula Sapiens Consortium, Stephen R Quake
bioRxiv 2024.12.03.626516; doi: https://doi.org/10.1101/2024.12.03.626516
