Domain expertise–agnostic feature selection for the analysis of breast cancer data*
作者：
Highlights：
• We propose a three-step wrapper method for the discovery of connected protein networks underlying particular molecular and cellular processes which characterize distinct behaviors in tumors in a manner complementary to the current PAM50-based breast cancer classification.
• Our method does not depend on a body of specialist knowledge and is complementary to current research on cancer biology.
• The protein clusters that showed top scoring modularities recapitulated many of the cellular phenotypes characteristic of cancer cells.
摘要
•We propose a three-step wrapper method for the discovery of connected protein networks underlying particular molecular and cellular processes which characterize distinct behaviors in tumors in a manner complementary to the current PAM50-based breast cancer classification.•Our method does not depend on a body of specialist knowledge and is complementary to current research on cancer biology.•The protein clusters that showed top scoring modularities recapitulated many of the cellular phenotypes characteristic of cancer cells.
论文关键词：Breast cancer,Clustering,Clustering performance evaluation,Dimensionality reduction,Feature selection,Proteomics,Unsupervised learning
论文评审过程：Received 10 December 2019, Revised 4 June 2020, Accepted 6 July 2020, Available online 16 July 2020, Version of Record 4 August 2020.
论文官网地址：https://doi.org/10.1016/j.artmed.2020.101928
