Adaptive multi-omics integration framework for breast cancer survival analysis

Breast cancer remains a major global health issue, requiring novel strategies for prognostic evaluation and therapeutic decision-making. In this study, we leverage multi-omics data from The Cancer Genome Atlas to obtain deeper insights into breast cancer biology. By integrating genomics, transcriptomics, and epigenomics, we aim to identify complex molecular signatures that drive breast cancer progression and impact patient survival. To optimize the integration and feature selection process within the multi-omics dataset, we have employed genetic programming. Genetic programming helps us to optimize multi-omics integration, enabling the identification of robust biomarkers and more accurate survival analysis. The proposed framework consists of three key components: data preprocessing, adaptive integration and feature selection via genetic programming, and model development. The experimental results indicate that the integrated multi-omics approach yields a concordance index (C-index) of 78.31 during 5 fold cross-validation on the training set and 67.94 on the test set. In conclusion, our study demonstrates the potential of adaptive multi-omics integration in improving breast cancer survival analysis. It also highlights the importance of considering the complex interplay between different molecular layers. Furthermore, this framework provides a flexible and scalable approach that can be extended to other cancer types, offering valuable insights into oncological processes.