Transformer-based AI technology improves early ovarian cancer diagnosis using cfDNA methylation markers

被引:1
|
作者
Li, Gen [1 ]
Zhang, Yongqiang [1 ]
Li, Kun [1 ]
Liu, Xiaohong [2 ,3 ,4 ,5 ,6 ]
Lu, Yaping [7 ]
Zhang, Zhenlin [2 ,3 ,4 ]
Liu, Zhihai [1 ]
Wu, Yong [8 ]
Liu, Fei [9 ]
Huang, Hong [1 ]
Yu, Meixing [1 ]
Yang, Zhao [1 ]
Zheng, Xiaoxue [1 ]
Guo, Chengbin [1 ,2 ]
Gao, Yuanxu [2 ,3 ,4 ,5 ]
Wang, Taorui [2 ,3 ,4 ]
Fok, Manson [2 ,3 ,4 ]
Lau, Johnson Yiu-Nam [10 ]
Shi, Kun [1 ]
Gu, Xiaoqiong [1 ]
Guo, Lingchuan [11 ]
Luo, Huiyan [12 ]
Zeng, Fanxin [13 ]
Zhang, Kang [2 ,3 ,4 ,5 ,14 ]
机构
[1] Guangzhou Women & Childrens Med Ctr, Guangzhou, Peoples R China
[2] Macau Univ Technol, Zhuhai Int Eye Ctr, Zhuhai, Peoples R China
[3] Macau Univ Technol, Zhuhai Peoples Hosp, Precis Med Ctr, Zhuhai, Peoples R China
[4] Macau Univ Technol, Affiliated Hosp 1, Fac Med, Zhuhai, Peoples R China
[5] Wenzhou Med Univ, Wenzhou Eye Hosp, Inst Adv Study Eye Hlth & Dis, Inst Clin Big Data, Wenzhou, Peoples R China
[6] UCL, Canc Inst, London WC1E 6BT, England
[7] Sinopharm Genom Technol Co Ltd, Sinopharm Med Lab Wuhan, Sinopharm Wuhan Precis Med Technol, Wuhan 430030, Peoples R China
[8] Jinan Univ, Guangzhou Overseas Chinese Hosp, Affiliated Hosp 1, Guangzhou, Peoples R China
[9] Chinese Acad Med Sci Peking Union Med Coll, Canc Hosp, Natl Canc Ctr, Natl Clin Res Ctr Canc, Beijing, Peoples R China
[10] Hong Kong Baptist Univ, Dept Biol, Hong Kong, Peoples R China
[11] Suzhou Univ, Affiliated Hosp 1, Dept Pathol, Suzhou, Peoples R China
[12] Sun Yat Sen Univ, Collaborat Innovat Ctr Canc Med, Canc Ctr, State Key Lab Oncol South China, Guangzhou, Peoples R China
[13] Dazhou Cent Hosp, Dept Clin Res Ctr, Dazhou, Peoples R China
[14] Guangzhou Natl Lab, Guangzhou, Peoples R China
基金
中国国家自然科学基金;
关键词
CIRCULATING TUMOR DNA; COLLABORATIVE TRIAL; ALGORITHM; MORTALITY; BENIGN; SCREEN; INDEX; CA125; WOMEN; ASSAY;
D O I
10.1016/j.xcrm.2024.101666
中图分类号
Q2 [细胞生物学];
学科分类号
071009 ; 090102 ;
摘要
Epithelial ovarian cancer (EOC) is the deadliest women's cancer and has a poor prognosis. Early detection is the key for improving survival (a 5-year survival rate in stage I/II is over 70% compared to that of 25% in stage III/IV) and can be achieved through methylation markers from circulating cell-free DNA (cfDNA) using a liquid biopsy. In this study, we first identify top 500 EOC markers differentiating EOC from healthy female controls from 3.3 million methylome-wide CpG sites and validated them in 1,800 independent cfDNA samples. We then utilize a pretrained AI transformer system called MethylBERT to develop an EOC diagnostic model which achieves 80% sensitivity and 95% specificity in early-stage EOC diagnosis. We next develop a simple digital droplet PCR (ddPCR) assay which archives good performance, facilitating early EOC detection.
引用
收藏
页数:18
相关论文
共 50 条
  • [31] Early diagnosis of ovarian cancer based on methylation profiles in peripheral blood cell-free DNA: a systematic review
    Terp, Simone Karlsson
    Stoico, Malene Pontoppidan
    Dybkaer, Karen
    Pedersen, Inge Sokilde
    CLINICAL EPIGENETICS, 2023, 15 (01)
  • [32] Breast cancer diagnosis through knowledge distillation of Swin transformer-based teacher-student models
    Kolla, Bhavannarayanna
    Venugopal, P.
    MACHINE LEARNING-SCIENCE AND TECHNOLOGY, 2023, 4 (04):
  • [33] Early detection of breast cancer in liquid biopsies using DNA methylation markers
    Neefs, I.
    Tjalma, W.
    Ibrahim, J.
    Croes, L.
    Peeters, M.
    Van Camp, G.
    de Beeck, K. Op
    BREAST, 2023, 68 : S30 - S31
  • [34] Multi-center Ovarian Tumor Classification Using Hierarchical Transformer-Based Multiple-Instance Learning
    Claessens, Cris H. B.
    Schultz, Eloy W. R.
    Koch, Anna
    Nies, Ingrid
    Hellstrom, Terese A. E.
    Nederend, Joost
    Niers-Stobbe, Ilse
    Bruining, Annemarie
    Piek, Jurgen M. J.
    De With, Peter H. N.
    van der Sommen, Fons
    CANCER PREVENTION, DETECTION, AND INTERVENTION, CAPTION 2024, 2025, 15199 : 3 - 13
  • [35] Investigation of the Ovarian and Prostate Cancer Peptidome for Candidate Early Detection Markers Using a Novel Nanoparticle Biomarker Capture Technology
    Claudia Fredolini
    Francesco Meani
    Alessandra Luchini
    Weidong Zhou
    Paul Russo
    Mark Ross
    Alexis Patanarut
    Davide Tamburro
    Guido Gambara
    David Ornstein
    Franco Odicino
    Monica Ragnoli
    Antonella Ravaggi
    Francesco Novelli
    Devis Collura
    Leonardo D’Urso
    Giovanni Muto
    Claudio Belluco
    Sergio Pecorelli
    Lance Liotta
    Emanuel F. Petricoin
    The AAPS Journal, 2010, 12 : 504 - 518
  • [36] Investigation of the Ovarian and Prostate Cancer Peptidome for Candidate Early Detection Markers Using a Novel Nanoparticle Biomarker Capture Technology
    Fredolini, Claudia
    Meani, Francesco
    Luchini, Alessandra
    Zhou, Weidong
    Russo, Paul
    Ross, Mark
    Patanarut, Alexis
    Tamburro, Davide
    Gambara, Guido
    Ornstein, David
    Odicino, Franco
    Ragnoli, Monica
    Ravaggi, Antonella
    Novelli, Francesco
    Collura, Devis
    D'Urso, Leonardo
    Muto, Giovanni
    Belluco, Claudio
    Pecorelli, Sergio
    Liotta, Lance
    Petricoin, Emanuel F., III
    AAPS JOURNAL, 2010, 12 (04): : 504 - 518
  • [37] Diagnosis of cancer at early stages based on the multiplex detection of tumor markers using metal nanoclusters
    Li, Dan
    Wang, Guannan
    Mei, Xifan
    Analyst, 2020, 145 (22): : 7150 - 7161
  • [38] Diagnosis of cancer at early stages based on the multiplex detection of tumor markers using metal nanoclusters
    Li, Dan
    Wang, Guannan
    Mei, Xifan
    ANALYST, 2020, 145 (22) : 7150 - 7161
  • [39] Detecting microsatellite instability in colorectal cancer using Transformer-based colonoscopy image classification and retrieval
    Lo, Chung-Ming
    Jiang, Jeng-Kai
    Lin, Chun-Chi
    PLOS ONE, 2024, 19 (01):
  • [40] MethylBERT enables read-level DNA methylation pattern identification and tumour deconvolution using a Transformer-based model
    Jeong, Yunhee
    Gerhaeuser, Clarissa
    Sauter, Guido
    Schlomm, Thorsten
    Rohr, Karl
    Lutsik, Pavlo
    NATURE COMMUNICATIONS, 2025, 16 (01)