Path2Omics Enhances Transcriptomic and Methylation Prediction Accuracy from Tumor Histopathology

Precision oncology is becoming increasingly integral to clinical practice, demonstrating notable improvements in treatment outcomes. Whereas molecular data provide comprehensive insights, obtaining such data remains costly and time-consuming. In this study, we developed Path2Omics, a deep learning framework that independently predicts gene expression and methylation from histopathology across 30 The Cancer Genome Atlas cancer types. Path2Omics comprised two components: a "formalin-fixed, paraffin-embedded (FFPE) model" trained on FFPE slides and a "fresh-frozen (FF) model" trained on FF slides. When evaluated on seven external datasets, the "FF model" outperformed the "FFPE model," even though six of the datasets consisted exclusively of FFPE slides. The "integrated model" combined predictions from both, achieving a 30% improvement over the FFPE model alone and robustly predicting approximately 4,400 genes (of 18,000). Importantly, the inferred gene expression closely matched actual values in predicting patient survival and treatment response. Overall, this study demonstrated the potential of Path2Omics to advance precision oncology using routine histopathology slides.

Path2Omics is an AI framework that predicts gene expression and methylation from histopathology slides, providing inferred values that are nearly as effective as actual values in predicting patient survival and treatment response. This article is part of a special series: Driving Cancer Discoveries with Computational Research, Data Science, and Machine Learning/AI .