Prognostic Biomarkers in Breast Cancer via Multi-Omics Clustering Analysis.

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Breast cancer (BC) is a heterogeneous disease, making prognosis and treatment challenging.
• Identifying reliable prognostic biomarkers is crucial for effective BC management.

Purpose of the Study:

• To identify key prognostic biomarkers for breast cancer using multi-omics data.
• To stratify breast cancer patients into subtypes with distinct survival outcomes.
• To explore the role of identified biomarkers in drug resistance.

Main Methods:

• Multi-omics clustering analysis using the Cancer Integration via MultIkernel LeaRning (CIMLR) method.
• Analysis of The Cancer Genome Atlas (TCGA) and METABRIC datasets.
• Validation using RNA sequencing data from therapy-resistant cell lines.

Main Results:

• Three genes (LMO1, PRAME, RSPO2) were identified as significantly associated with poor prognosis in BC.
• Patient stratification based on these genes revealed distinct subtypes with different overall survival (OS).
• LMO1 and PRAME were upregulated in patients with worse prognosis and therapy-resistant cells, suggesting a role in drug resistance.

Conclusions:

• LMO1 and PRAME are potential biomarkers for identifying high-risk breast cancer patients.
• These biomarkers can inform targeted treatment strategies and personalized therapeutic approaches.
• The study provides insights into the multi-omics landscape of breast cancer.