Optimizing Variable Selection in Multi-Omics Datasets: A Focus on Exclusive Lasso

Multi-omics datasets pose significant challenges due to their structured nature, where highly correlated variables are grouped within a complex, high-dimensional framework. Traditional Lasso methods encounter limitations in handling correlated features within these groups effectively. To address this issue, we propose using Exclusive Lasso, focusing on inducing sparsity at the intra-group level. Additionally, we introduce an efficient algorithm for solving the related optimization problem. By prioritizing feature selection robustness within correlated group structures, our proposed methodology offers a promising solution to the challenges inherent in analyzing biological datasets. This advancement enhances our ability to extract meaningful insights from multi-omics data, thus facilitating deeper understanding and exploration of complex biological systems.