Deep learning integrates histopathology and proteogenomics at a pan-cancer level

• DOI: 10.1016/j.xcrm.2023.101173
• Publication Date: 2023-08-14T14:26:55Z
• Authors: Joshua M. Wang; Runyu Hong; Elizabeth G. Demicco; Jimin Tan; Rossana Lazcano; Andre L. Moreira; Yize Li; Anna Calinawan; Narges Razavian; Tobias Schraink; Michael A. Gillette; Gilbert S. Omenn; Eunkyung An; Henry Rodriguez; Aristotelis Tsirigos; Kelly V. Ruggles; Li Ding; Ana I. Robles; D.R. Mani; Karin D. Rodland; Alexander J. Lazar; Wenke Liu; David Fenyö; François Aguet; Yo Akiyama; Shankara Anand; Meenakshi Anurag; Özgün Babur; Jasmin Bavarva; Chet Birger; Michael J. Birrer; Lewis C. Cantley; Song Cao; Steven A. Carr; Michele Ceccarelli; Daniel W. Chan; Arul M. Chinnaiyan; Hanbyul Cho; Shrabanti Chowdhury; Marcin P. Cieslik; Karl R. Clauser; Antonio Colaprico; Daniel Cui Zhou; Felipe da Veiga L...; Corbin Day; Saravana M. Dhana...; Marcin J. Domagalski; Yongchao Dou; Brian J. Druker; Nathan Edwards; Matthew J. Ellis; Myvizhi Esai Selvan; Steven M. Foltz; Alicia Francis; Yifat Geffen; Gad Getz; Tania J. Gonzalez...; Sara J.C. Gosline; Zeynep H. Gümüş; David I. Heiman; Tara Hiltke; Galen Hostetter; Yingwei Hu; Chen Huang; Emily Huntsman; Antonio Iavarone; Eric J. Jaehnig; Scott D. Jewell; Jiayi Ji; Wen Jiang; Jared L. Johnson; Lizabeth Katsnelson; Karen A. Ketchum; Iga Kolodziejczak; Karsten Krug; Chandan Kumar-Sinha; Jonathan T. Lei; Wen-Wei Liang; Yuxing Liao; Caleb M. Lindgren; Tao Liu; Weiping Ma; Fernanda Martins ...; Wilson McKerrow; Mehdi Mesri; Alexey I. Nesvizh...; Chelsea J. Newton; Robert Oldroyd; Amanda G. Paulovich; Samuel H. Payne; Francesca Petralia; Pietro Pugliese; Boris Reva; Dmitry Rykunov; Shankha Satpathy; Sara R. Savage; Eric E. Schadt; Michael Schnaubelt; Stephan Schürer; Zhiao Shi; Richard D. Smith; Xiaoyu Song; Yizhe Song; Vasileios Stathias; Erik P. Storrs; Nadezhda V. Terek...; Ratna R. Thangudu; Mathangi Thiagarajan; Nicole Tignor; Liang-Bo Wang; Pei Wang; Ying Wang; Bo Wen; Maciej Wiznerowicz; Yige Wu; Matthew A. Wyczal...; Lijun Yao; Tomer M. Yaron; Xinpei Yi; Bing Zhang; Hui Zhang; Qing Zhang; Xu Zhang; Zhen Zhang
• Keywords: Proteogenomics; Interpretability; Predictive power

We introduce a pioneering approach that integrates pathology imaging with transcriptomics and proteomics to identify predictive histology features associated critical clinical outcomes in cancer. utilize 2,755 H&E-stained histopathological slides from 657 patients across 6 cancer types CPTAC. Our models effectively recapitulate distinctions readily made by human pathologists: tumor vs. normal (AUROC = 0.995) tissue-of-origin 0.979). further investigate power on tasks not normally performed H&E alone, including TP53 prediction pathologic stage. Importantly, we describe morphologies previously utilized setting. The incorporation of identifies pathway-level signatures cellular processes driving features. Model generalizability interpretability is confirmed using TCGA. propose classification system for these tasks, suggest potential applications this integrated machine learning approach. A publicly available web-based platform implements models.