Research Paper Volume 15, Issue 22 pp 12890—12906
TJP3 promotes T cell immunity escape and chemoresistance in breast cancer: a comprehensive analysis of anoikis-based prognosis prediction and drug sensitivity stratification
Figure 1. Research Process. (1) Identification of ARG subgroup: gene expression profiles are collected from TCGA and GEO databases, and ARG list is collected from GeneCard. 44 ARGs are selected into consensus cluster to identify subgroups. (2) Features in subgroups: Pathway score and immune cell infiltration are calculated by ssGSEA and CIBERSORT, respectively. (3) AI-based drug sensitivity stratification: 5 types of machine learning algorithms are applied, and XGBoost displays best results in identification of ARG subgroups. (4) TCGA (training) and GEO (testing) cohorts are used to construct prognosis prediction model. (5) Hubgenes are selected to perform in vitro experiments.