Discovery and Validation of Key Biomarkers based on Machine Learning and Immune Infiltrates in Ovarian Cancer

Linlin Zhang, Mingming Yu, Xuehua Bi, Guanglei Yu, Kai Zhao

2022

Abstract

Ovarian cancer (OC) is the deadliest gynecological malignancy which survival rate mainly depends on early detection. Our purpose was to search for potential OC diagnostic markers and to examine the role of immune cell infiltration in its disease process. OC expression profiles were extracted from Gene Expression Omnibus (GEO) and differentially expressed genes (DEGs) were identified with the limma R package and subjected to functional correlation analysis. We used Hilbert-Schmidt Independence Criterion Least Absolute Shrinkage and Selection Operator (HSIC-Lasso), Support Vector Machine-Recursive Feature Elimination (SVM-RFE) algorithms and Minimum Redundancy Maximum Relevance (mRMR) to select gene features and chose the random forest (RF) algorithm as the classifier to validate the results of gene selection. Finally, we utilized CIBERSORT to bulk gene ex-pression profiles of OC for quantifying 22 subsets of immune cells. Subsequently, we analysed the correlation between diagnostic markers and infiltrating immune cells. ABCA8, IGFBP2 and REEP1 were identified as diagnostic markers for OC in this study (AUC=0.96), and a total of 380 DEGs were identified. Immune cell infiltration analysis showed that plasma cells, CD8 T cells and activated memory CD4 T cells may be involved in the occurrence and development of OC. In addition, ABCA8 was positively correlated with neutrophils, monocytes, activated NK cells while negatively correlated with activated CD4 memory T cells, naïve B cells and macrophages M1. IGFBP2 was positively correlated with macrophages M1 while negatively correlated with monocytes and neutrophils. REEP1 was positively correlated with neutrophils, monocytes, macrophages M2, activated NK cells and plasma cells while negatively correlated with resting NK cells, activated CD4 memory T cells and CD8 T cells. In conclusion, ABCA8, IGFBP2 and REEP1 can be used as diagnostic markers of OC, and immune cell infiltration plays a crucial role in the occurrence and progression of OC.

Paper Citation

in Harvard Style

Zhang L., Yu M., Bi X., Yu G. and Zhao K. (2022). Discovery and Validation of Key Biomarkers based on Machine Learning and Immune Infiltrates in Ovarian Cancer. In Proceedings of the 4th International Conference on Biotechnology and Biomedicine - Volume 1: ICBB; ISBN 978-989-758-637-8, SciTePress, pages 253-265. DOI: 10.5220/0012019200003633

in Bibtex Style

@conference{icbb22,
author={Linlin Zhang and Mingming Yu and Xuehua Bi and Guanglei Yu and Kai Zhao},
title={Discovery and Validation of Key Biomarkers based on Machine Learning and Immune Infiltrates in Ovarian Cancer},
booktitle={Proceedings of the 4th International Conference on Biotechnology and Biomedicine - Volume 1: ICBB},
year={2022},
pages={253-265},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012019200003633},
isbn={978-989-758-637-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 4th International Conference on Biotechnology and Biomedicine - Volume 1: ICBB
TI - Discovery and Validation of Key Biomarkers based on Machine Learning and Immune Infiltrates in Ovarian Cancer
SN - 978-989-758-637-8
AU - Zhang L.
AU - Yu M.
AU - Bi X.
AU - Yu G.
AU - Zhao K.
PY - 2022
SP - 253
EP - 265
DO - 10.5220/0012019200003633
PB - SciTePress