Universidad de Navarra, Tecnun Escuela de Ingeniería, San Sebastián, Spain.
Universidad de Navarra, CIMA Centro de Investigación de Medicina Aplicada, Pamplona, Spain.
PLoS Comput Biol. 2022 May 31;18(5):e1010180. doi: 10.1371/journal.pcbi.1010180. eCollection 2022 May.
With the frenetic growth of high-dimensional datasets in different biomedical domains, there is an urgent need to develop predictive methods able to deal with this complexity. Feature selection is a relevant strategy in machine learning to address this challenge. We introduce a novel feature selection algorithm for linear regression called BOSO (Bilevel Optimization Selector Operator). We conducted a benchmark of BOSO with key algorithms in the literature, finding a superior accuracy for feature selection in high-dimensional datasets. Proof-of-concept of BOSO for predicting drug sensitivity in cancer is presented. A detailed analysis is carried out for methotrexate, a well-studied drug targeting cancer metabolism.
随着不同生物医学领域高维数据集的飞速增长，我们迫切需要开发能够应对这种复杂性的预测方法。特征选择是机器学习中应对这一挑战的一种相关策略。我们引入了一种新的线性回归特征选择算法，称为 BOSO（双层优化选择算子）。我们对 BOSO 与文献中的关键算法进行了基准测试，发现其在高维数据集中进行特征选择的准确性更高。我们还展示了 BOSO 用于预测癌症药物敏感性的概念验证。我们对甲氨蝶呤（一种针对癌症代谢的研究充分的药物）进行了详细分析。
