Multiple Omics Data Integration to Identify Long Noncoding RNA Responsible for Breast Cancer-Related Mortality

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Long non-coding RNAs (lncRNAs) are implicated in cancer development.
• Accurate prediction of patient survival is crucial for effective breast cancer treatment.

Purpose of the Study:

• To develop an integrative modeling framework to predict breast cancer patient survival.
• To integrate DNA copy number variation (CNV), lncRNA expression, and target protein expression data.
• To evaluate the predictive performance of a novel 3-stage model.

Main Methods:

• An integrative modeling framework combining mechanical and clinical models was developed.
• The model integrated CNV, lncRNA expression (e.g., HOTAIR, MALAT1), and downstream target protein expression.
• Data from The Cancer Genome Atlas (TCGA) database was utilized for model training and validation.

Main Results:

• The proposed 3-stage integrative model demonstrated lower predicted mean square error and integrated Brier score (IBS) compared to a 2-step model.
• The model incorporating target protein information showed superior predictive ability.
• Specific lncRNAs like HOTAIR and MALAT1 were analyzed within the framework.

Conclusions:

• The integrative 3-stage model offers enhanced predictive accuracy for breast cancer patient survival.
• Integrating multi-omics data, including CNV, lncRNA, and protein expression, improves survival prediction.
• This framework provides a valuable tool for personalized breast cancer prognosis.