Identification of Key Genes and Key Pathways in Breast Cancer Based on Machine Learning

Abstract

Background Breast cancer is one of the most common malignant tumors among women worldwide. This study aimed to screen key genes and pathways for breast cancer diagnosis and treatment. Material/Methods We obtained public data from the NCBI GEO database. The data were divided into a control group (normal breast tissue) and a treatment group (breast cancer tissue). We screened 32 differentially expressed genes (DEGs) between normal breast and cancerous tissues and used GO analysis and GSEA to identify the key pathways. We then combined LASSO and SVM-RFE analyses to screen key genes, and used CIBERSORT to obtain the proportion of 22 types of immune cells. The relationships between key genes and immune-infiltrating cells were further explored. Results We screened 32 DEGs from the 2 groups, including 27 downregulated genes and 5 upregulated genes. GO analysis indicated that the DEGs were mainly correlated with collagen-containing extracellular matrix (ECM), Wnt signaling pathway, and glycosaminoglycan binding. GSEA indicated that the treatment group was correlated with chromosome segregation and cell cycle while the control group was correlated with cornification, intermediate filament, and nuclear transcription. Through machine learning, SYNM , TGFBR3 , and COL10A1 were screened as key genes. Numbers of CD8 T cells, gamma delta T cells, and M1 macrophages were significantly higher, while monocytes and follicular helper-T cells were significantly lower in the treatment group. The downregulated genes, SYNM and TGFBR3 , were positively correlated with CD8 T cells and monocytes, but were negatively correlated with gamma delta T cells and M1 macrophages. The upregulated gene, COL10A1 , was positively correlated with gamma delta T cells and M1 macrophages, and was negatively correlated with CD8 T cells, monocytes, and follicular helper-T cells. Conclusions SYNM , TGFBR3 , and COL10A1 are diagnostic genes of breast cancer. They affect breast cancer cells by modulating immune-infiltrating cells.

Identification of Novel Diagnostic and Prognostic Gene Signature Biomarkers for Breast Cancer Using Artificial Intelligence and Machine Learning Assisted Transcriptomics Analysis

“…First, to verify the reliability of our method of screening for biomarkers, we confirmed our finding of DEGs, pathways, and gene ontologies using literature mining and verification. Matching our results with previous findings was good evidence that they are indeed involved in the development and progression of BC [ 42 , 43 , 44 , 45 , 46 ].…”

Unveiling Collagen’s Role in Breast Cancer: Insights into Expression Patterns, Functions and Clinical Implications

“…Currently, ongoing research has revealed its impact on breast cancer cells by modulating immune-infiltrating cells. 62 It is positively associated with immune cell infiltration, including B cell, CD8 + T cell, CD4 + T cell, macrophage, neutrophil, and dendritic cell. 63 COL10A1’s direct interaction with Prolyl 4-hydroxylase beta polypeptide (P4HB) has been identified, leading to the upregulation of P4HB expression, thereby promoting malignant progression in breast cancer.…”