Integrating cell-free DNA methylation of SEPT9 and SFRP2 into a machine learning model for early diagnosis of HCC.

IF 2.1 4区医学 Q3 MEDICINE, RESEARCH & EXPERIMENTAL

Biomarkers in medicine Pub Date : 2025-08-01 Epub Date: 2025-08-04 DOI:10.1080/17520363.2025.2541574

Dong Wang, Zhihao Dai, Minghua Bai, Dong Liu, Yanru Feng, Quanquan Sun, Tong Zhang, Liang Han, Rui Wang, Ji Zhu, Weifeng Hong, Weiwei Li

{"title":"Integrating cell-free DNA methylation of SEPT9 and SFRP2 into a machine learning model for early diagnosis of HCC.","authors":"Dong Wang, Zhihao Dai, Minghua Bai, Dong Liu, Yanru Feng, Quanquan Sun, Tong Zhang, Liang Han, Rui Wang, Ji Zhu, Weifeng Hong, Weiwei Li","doi":"10.1080/17520363.2025.2541574","DOIUrl":null,"url":null,"abstract":"Background: Hepatocellular carcinoma (HCC), a primary contributor to cancer-associated mortality, necessitates enhanced early detection. This study evaluated machine learning models that merge methylated SEPTIN9 (SEPT9) and secreted frizzled-related protein 2 (SFRP2) within circulating cell-free DNA (cfDNA) to detect HCC.Methods: A cohort of 165 healthy volunteers, 24 precancerous patients of HCC and 112 HCC patients were divided into training and validation sets. Methylated SEPT9 and SFRP2 (mSEPT9/mSFRP2) were detected using real-time PCR. Based on those methylation biomarkers and/or conventional biomarkers (CEA, AFP, CA125, and CA19-9), six machine learning algorithms, including Random Forest (RF), were employed to establish models for the training set. Models were evaluated for area under the ROC curve (AUC), sensitivity, and specificity, and subsequently validated in the validation set.Results: The RF model outperformed other models. In training, it achieved an AUC of 0.834 (95% CI: 0.745-0.923), exhibiting 69.3% sensitivity and 80.6% specificity for the methylation-specific signature group (mSS group: mSEPT9/mSFRP2). In validation, the RF model for the mSS group showed an AUC of 0.865 (95% CI: 0.811-0.946), with 85.4% sensitivity and 71.4% specificity.Conclusions: The RF-based model integrating mSEPT9/mSFRP2 in cfDNA can be a promising approach for HCC diagnosis.","PeriodicalId":9182,"journal":{"name":"Biomarkers in medicine","volume":" ","pages":"737-745"},"PeriodicalIF":2.1000,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12416165/pdf/","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Biomarkers in medicine","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1080/17520363.2025.2541574","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2025/8/4 0:00:00","PubModel":"Epub","JCR":"Q3","JCRName":"MEDICINE, RESEARCH & EXPERIMENTAL","Score":null,"Total":0}

引用次数: 0

Abstract

Background: Hepatocellular carcinoma (HCC), a primary contributor to cancer-associated mortality, necessitates enhanced early detection. This study evaluated machine learning models that merge methylated SEPTIN9 (SEPT9) and secreted frizzled-related protein 2 (SFRP2) within circulating cell-free DNA (cfDNA) to detect HCC.

Methods: A cohort of 165 healthy volunteers, 24 precancerous patients of HCC and 112 HCC patients were divided into training and validation sets. Methylated SEPT9 and SFRP2 (mSEPT9/mSFRP2) were detected using real-time PCR. Based on those methylation biomarkers and/or conventional biomarkers (CEA, AFP, CA125, and CA19-9), six machine learning algorithms, including Random Forest (RF), were employed to establish models for the training set. Models were evaluated for area under the ROC curve (AUC), sensitivity, and specificity, and subsequently validated in the validation set.

Results: The RF model outperformed other models. In training, it achieved an AUC of 0.834 (95% CI: 0.745-0.923), exhibiting 69.3% sensitivity and 80.6% specificity for the methylation-specific signature group (mSS group: mSEPT9/mSFRP2). In validation, the RF model for the mSS group showed an AUC of 0.865 (95% CI: 0.811-0.946), with 85.4% sensitivity and 71.4% specificity.

Conclusions: The RF-based model integrating mSEPT9/mSFRP2 in cfDNA can be a promising approach for HCC diagnosis.

查看原文本刊更多论文

将SEPT9和SFRP2的游离DNA甲基化整合到HCC早期诊断的机器学习模型中。

背景：肝细胞癌（HCC）是癌症相关死亡的主要原因之一，需要加强早期检测。该研究评估了在循环无细胞DNA （cfDNA）中合并甲基化SEPTIN9 （SEPT9）和分泌卷曲相关蛋白2 （SFRP2）的机器学习模型，以检测HCC。方法：将165名健康志愿者、24名肝癌前患者和112名肝癌患者分为训练组和验证组。实时荧光定量PCR检测甲基化的SEPT9和SFRP2 （mSEPT9/mSFRP2）。基于甲基化生物标记物和/或常规生物标记物（CEA、AFP、CA125和CA19-9），采用随机森林（Random Forest， RF）等6种机器学习算法建立训练集模型。评估模型的ROC曲线下面积（AUC）、敏感性和特异性，并随后在验证集中进行验证。结果：RF模型优于其他模型。在训练中，它的AUC为0.834 (95% CI: 0.745-0.923)，对甲基化特异性特征组（mSS组：mSEPT9/mSFRP2）的敏感性为69.3%，特异性为80.6%。在验证中，mSS组的RF模型的AUC为0.865 (95% CI: 0.811-0.946)，敏感性为85.4%，特异性为71.4%。结论：在cfDNA中整合mSEPT9/mSFRP2的基于rf的模型可能是一种很有前景的HCC诊断方法。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Biomarkers in medicine 医学-医学：研究与实验

CiteScore

3.80

自引率

4.50%

发文量

审稿时长

6-12 weeks

期刊介绍： Biomarkers are physical, functional or biochemical indicators of physiological or disease processes. These key indicators can provide vital information in determining disease prognosis, in predicting of response to therapies, adverse events and drug interactions, and in establishing baseline risk. The explosion of interest in biomarker research is driving the development of new predictive, diagnostic and prognostic products in modern medical practice, and biomarkers are also playing an increasingly important role in the discovery and development of new drugs. For the full utility of biomarkers to be realized, we require greater understanding of disease mechanisms, and the interplay between disease mechanisms, therapeutic interventions and the proposed biomarkers. However, in attempting to evaluate the pros and cons of biomarkers systematically, we are moving into new, challenging territory. Biomarkers in Medicine (ISSN 1752-0363) is a peer-reviewed, rapid publication journal delivering commentary and analysis on the advances in our understanding of biomarkers and their potential and actual applications in medicine. The journal facilitates translation of our research knowledge into the clinic to increase the effectiveness of medical practice. As the scientific rationale and regulatory acceptance for biomarkers in medicine and in drug development become more fully established, Biomarkers in Medicine provides the platform for all players in this increasingly vital area to communicate and debate all issues relating to the potential utility and applications. Each issue includes a diversity of content to provide rounded coverage for the research professional. Articles include Guest Editorials, Interviews, Reviews, Research Articles, Perspectives, Priority Paper Evaluations, Special Reports, Case Reports, Conference Reports and Company Profiles. Review coverage is divided into themed sections according to area of therapeutic utility with some issues including themed sections on an area of topical interest. Biomarkers in Medicine provides a platform for commentary and debate for all professionals with an interest in the identification of biomarkers, elucidation of their role and formalization and approval of their application in modern medicine. The audience for Biomarkers in Medicine includes academic and industrial researchers, clinicians, pathologists, clinical chemists and regulatory professionals.