Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2506.03373 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918045366091776 |
|---|---|
| author | Shaban, Muhammad Chang, Yuzhou Qiu, Huaying Yeo, Yao Yu Song, Andrew H. Jaume, Guillaume Wang, Yuchen Weishaupt, Luca L. Ding, Tong Vaidya, Anurag Lamane, Abdallah Shao, Daniel Zidane, Mohammed Bai, Yunhao McCallum, Paige Luo, Shuli Wu, Wenrui Wang, Yang Cramer, Precious Chan, Chi Ngai Stephan, Pierre Schaffenrath, Johanna Lee, Jia Le Michel, Hendrik A. Tian, Caiwei Almagro-Perez, Cristina Wagner, Sophia J. Sahai, Sharifa Lu, Ming Y. Chen, Richard J. Zhang, Andrew Gonzales, Mark Edward M. Makky, Ahmad Lee, Jia-Ying Joey Cheng, Hao Ahmar, Nourhan El Matar, Sayed Haist, Maximilian Phillips, Darci Tan, Yuqi Nolan, Garry P. Burack, W. Richard Estes, Jacob D. Liu, Jonathan T. C. Choueiri, Toni K Agarwal, Neeraj Barry, Marc Rodig, Scott J. Le, Long Phi Gerber, Georg Schürch, Christian M. Theis, Fabian J. Kim, Youn H Yeong, Joe Signoretti, Sabina Howitt, Brooke E. Loo, Lit-Hsin Ma, Qin Jiang, Sizun Mahmood, Faisal |
| author_facet | Shaban, Muhammad Chang, Yuzhou Qiu, Huaying Yeo, Yao Yu Song, Andrew H. Jaume, Guillaume Wang, Yuchen Weishaupt, Luca L. Ding, Tong Vaidya, Anurag Lamane, Abdallah Shao, Daniel Zidane, Mohammed Bai, Yunhao McCallum, Paige Luo, Shuli Wu, Wenrui Wang, Yang Cramer, Precious Chan, Chi Ngai Stephan, Pierre Schaffenrath, Johanna Lee, Jia Le Michel, Hendrik A. Tian, Caiwei Almagro-Perez, Cristina Wagner, Sophia J. Sahai, Sharifa Lu, Ming Y. Chen, Richard J. Zhang, Andrew Gonzales, Mark Edward M. Makky, Ahmad Lee, Jia-Ying Joey Cheng, Hao Ahmar, Nourhan El Matar, Sayed Haist, Maximilian Phillips, Darci Tan, Yuqi Nolan, Garry P. Burack, W. Richard Estes, Jacob D. Liu, Jonathan T. C. Choueiri, Toni K Agarwal, Neeraj Barry, Marc Rodig, Scott J. Le, Long Phi Gerber, Georg Schürch, Christian M. Theis, Fabian J. Kim, Youn H Yeong, Joe Signoretti, Sabina Howitt, Brooke E. Loo, Lit-Hsin Ma, Qin Jiang, Sizun Mahmood, Faisal |
| contents | Foundation models have begun to transform image analysis by acting as pretrained generalist backbones that can be adapted to many tasks even when post-training data are limited, yet their impact on spatial proteomics, imaging that maps proteins at single-cell resolution, remains limited. Here, we introduce KRONOS, a foundation model built for spatial proteomics. KRONOS was trained in a self-supervised manner on over 47 million image patches covering 175 protein markers, 16 tissue types, and 8 fluorescence-based imaging platforms. We introduce key architectural adaptations to address the high-dimensional, multi-channel, and heterogeneous nature of multiplex imaging. We demonstrate that KRONOS learns biologically meaningful representations across multiple scales, ranging from cellular and microenvironment to tissue levels, enabling it to address diverse downstream tasks, including cell phenotyping, region classification, and patient stratification. Evaluated across 11 independent cohorts, KRONOS achieves state-of-the-art performance across cell phenotyping, treatment response prediction, and retrieval tasks, and is highly data-efficient. KRONOS also introduces the paradigm of segmentation-free patch-level processing for efficient and scalable spatial proteomics analysis, allowing cross-institutional comparisons, and as an image reverse search engine for spatial patterns. Together, these results position KRONOS as a flexible and scalable tool for spatial proteomics. The model is publicly accessible at https://github.com/mahmoodlab/KRONOS. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_03373 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | A Foundation Model for Spatial Proteomics Shaban, Muhammad Chang, Yuzhou Qiu, Huaying Yeo, Yao Yu Song, Andrew H. Jaume, Guillaume Wang, Yuchen Weishaupt, Luca L. Ding, Tong Vaidya, Anurag Lamane, Abdallah Shao, Daniel Zidane, Mohammed Bai, Yunhao McCallum, Paige Luo, Shuli Wu, Wenrui Wang, Yang Cramer, Precious Chan, Chi Ngai Stephan, Pierre Schaffenrath, Johanna Lee, Jia Le Michel, Hendrik A. Tian, Caiwei Almagro-Perez, Cristina Wagner, Sophia J. Sahai, Sharifa Lu, Ming Y. Chen, Richard J. Zhang, Andrew Gonzales, Mark Edward M. Makky, Ahmad Lee, Jia-Ying Joey Cheng, Hao Ahmar, Nourhan El Matar, Sayed Haist, Maximilian Phillips, Darci Tan, Yuqi Nolan, Garry P. Burack, W. Richard Estes, Jacob D. Liu, Jonathan T. C. Choueiri, Toni K Agarwal, Neeraj Barry, Marc Rodig, Scott J. Le, Long Phi Gerber, Georg Schürch, Christian M. Theis, Fabian J. Kim, Youn H Yeong, Joe Signoretti, Sabina Howitt, Brooke E. Loo, Lit-Hsin Ma, Qin Jiang, Sizun Mahmood, Faisal Computer Vision and Pattern Recognition Artificial Intelligence Foundation models have begun to transform image analysis by acting as pretrained generalist backbones that can be adapted to many tasks even when post-training data are limited, yet their impact on spatial proteomics, imaging that maps proteins at single-cell resolution, remains limited. Here, we introduce KRONOS, a foundation model built for spatial proteomics. KRONOS was trained in a self-supervised manner on over 47 million image patches covering 175 protein markers, 16 tissue types, and 8 fluorescence-based imaging platforms. We introduce key architectural adaptations to address the high-dimensional, multi-channel, and heterogeneous nature of multiplex imaging. We demonstrate that KRONOS learns biologically meaningful representations across multiple scales, ranging from cellular and microenvironment to tissue levels, enabling it to address diverse downstream tasks, including cell phenotyping, region classification, and patient stratification. Evaluated across 11 independent cohorts, KRONOS achieves state-of-the-art performance across cell phenotyping, treatment response prediction, and retrieval tasks, and is highly data-efficient. KRONOS also introduces the paradigm of segmentation-free patch-level processing for efficient and scalable spatial proteomics analysis, allowing cross-institutional comparisons, and as an image reverse search engine for spatial patterns. Together, these results position KRONOS as a flexible and scalable tool for spatial proteomics. The model is publicly accessible at https://github.com/mahmoodlab/KRONOS. |
| title | A Foundation Model for Spatial Proteomics |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2506.03373 |