Multimodal alignment improves generalizability of genomic biomarker prediction in computational pathology
- 1Microsoft Research, Cambridge, MA 02139, USA
Abstract
Computational pathology models that use digitized histopathology whole-slide images have the potential to become a cost-effective and scalable alternative to molecular assays for the prediction of genomic biomarkers, a key task in precision oncology. However, as new genomic biomarkers are discovered or quantified, large, labeled datasets must be prospectively collected to train new models. To address this challenge, we developed multimodal alignment for biomarker learning and generalization (MARBLE), a multimodal contrastive pretraining strategy that integrates structured biomarker knowledge into representation learning of histopathology images. MARBLE aligns histopathology-derived representations with representations of genomic biomarkers generated by a large language model (LLM) and a protein language model (PLM). This biologically informed alignment enables data-efficient generalization to novel, out-of-distribution biomarkers. Using the MSK-IMPACT cohort of over 40,000 patients across multiple biomarker panel versions, we design experiments grounded in real-world data to demonstrate the value of our proposed approach.
Related
- Systematic comparison of affinity plasma proteomic technologies across multiple cohorts reveals generalizability of proteomic prediction
- Machine learning prediction of rheumatoid arthritis using the gut microbiome shows limited cross-cohort generalizability
- Computational design and molecular dynamics for developing a multistage serodiagnostic biomarker for tuberculosis
- Explainable computational framework for subtype classification and biomarker identification in head and neck squamous cell carcinoma using machine learning
- Katalyst: knowledge-guided semantic alignment via contrastive learning for enzyme turnover prediction
BibTeX
@article{redekop2026multimodal,
title = {{Multimodal alignment improves generalizability of genomic biomarker prediction in computational pathology}},
author = {Ekaterina Redekop and Eric Zimmermann and Ava P. Amini and Alex X. Lu and Neil Tenenholtz and James Hall and Lorin Crawford and Kristen A. Severson},
journal = {Cell Reports Methods},
year = {2026},
month = sep,
doi = {10.1016/j.crmeth.2026.101578},
url = {https://doi.org/10.1016/j.crmeth.2026.101578},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - Multimodal alignment improves generalizability of genomic biomarker prediction in computational pathology
AU - Ekaterina Redekop
AU - Eric Zimmermann
AU - Ava P. Amini
AU - Alex X. Lu
AU - Neil Tenenholtz
AU - James Hall
AU - Lorin Crawford
AU - Kristen A. Severson
PY - 2026
DA - 2026/09/08
JO - Cell Reports Methods
DO - 10.1016/j.crmeth.2026.101578
UR - https://doi.org/10.1016/j.crmeth.2026.101578
AB - Computational pathology models that use digitized histopathology whole-slide images have the potential to become a cost-effective and scalable alternative to molecular assays for the prediction of genomic biomarkers, a key task in precision oncology. However, as new genomic biomarkers are discovered or quantified, large, labeled datasets must be prospectively collected to train new models. To address this challenge, we developed multimodal alignment for biomarker learning and generalization (MARBLE), a multimodal contrastive pretraining strategy that integrates structured biomarker knowledge into representation learning of histopathology images. MARBLE aligns histopathology-derived representations with representations of genomic biomarkers generated by a large language model (LLM) and a protein language model (PLM). This biologically informed alignment enables data-efficient generalization to novel, out-of-distribution biomarkers. Using the MSK-IMPACT cohort of over 40,000 patients across multiple biomarker panel versions, we design experiments grounded in real-world data to demonstrate the value of our proposed approach.
SN - 2667-2375
ER -