Related Experiment Video
Updated: Jun 15, 2025
Profiling of Estrogen-regulated MicroRNAs in Breast Cancer Cells
Published on: February 21, 2014
A Deep Differential Analysis in Four Subtypes of Breast Cancer Based on Regulations of miRNA-mRNA
Tao Huang1, Ling Guo1, Weiyuan Ma2
1Department of Electrical Engineering, Northwest Minzu University, Lanzhou, China.
Abstract:
Breast cancer is a highly heterogeneous disease and it is generally divided into four subtypes in clinical practice. Common differentially expressed genes are always ignored. In fact, the regulatory associations of common differentially expressed genes exhibit significant differences among the four subtypes of breast cancer. A deep differential analysis in four subtype of breast cancer is proposed in this paper. The common differentially expressed genes among four subtypes of breast cancer are mainly considered. The miRNA-mRNA regulatory network is constructed as a bipartite network and the regulations of miRNA-mRNA for each subtype of breast cancer are predicted. The common differentially expressed genes for four subtypes of breast cancer are obtained. Breast cancer is classified into four subtypes by using Prediction Analysis of Microarray 50. The method of EdgeR is employed to obtain the common differentially expressed genes. A background network is designed by the common differentially expressed genes. MiRNA-mRNA bipartite network is constructed by the background network. A method of weighted similarity information (WSI) is proposed. Global similarity information of miRNA and mRNA are obtained by the WSI, respectively. The regulations of miRNA-mRNA in four subtypes of breast cancer are predicted by integrating the MiRNA-mRNA bipartite network and the global similarity information of miRNA and mRNA. In 5-fold cross-validation, this method performs well across the four subtypes of breast cancer. In addition, the predicted regulations of miRNA-mRNA have 85% ratio in the miRWalk2.0 database. This represents a 30% improvement over traditional methods.
Insights
This study reveals significant differences in gene regulation across breast cancer subtypes. A novel method accurately predicts microRNA-messenger RNA interactions, improving upon traditional approaches.
Area of Science:
- Oncology
- Bioinformatics
- Genomics
Background:
- Breast cancer is a heterogeneous disease classified into four subtypes.
- Commonly differentially expressed genes are often overlooked, despite varying regulatory associations among subtypes.
- Understanding subtype-specific gene regulation is crucial for targeted therapies.
Purpose of the Study:
- To perform a deep differential analysis of gene expression across four breast cancer subtypes.
- To investigate the regulatory associations of common differentially expressed genes.
- To develop a novel method for predicting microRNA-mRNA regulatory networks specific to each breast cancer subtype.
Main Methods:
- Classified breast cancer subtypes using Prediction Analysis of Microarray 50.
- Identified common differentially expressed genes using the EdgeR method.
- Constructed miRNA-mRNA bipartite networks and predicted regulations using a weighted similarity information (WSI) method.
Main Results:
- Developed a novel method for predicting miRNA-mRNA regulations in breast cancer subtypes.
- Achieved an 85% accuracy rate for predicted regulations against the miRWalk2.0 database.
- Demonstrated a 30% improvement in prediction accuracy compared to traditional methods.
Conclusions:
- The proposed deep differential analysis and WSI method effectively capture subtype-specific miRNA-mRNA regulatory differences.
- This approach offers a significant advancement in understanding breast cancer heterogeneity and developing targeted strategies.
- The findings provide a valuable resource for future breast cancer research and therapeutic development.
Related Concept Videos
01:22MicroRNAs
01:23Regulation of Expression at Multiple Steps