fig5

Figure 5. Multi-omics identification of immune subtypes in LUAD. (A) Multi-omics clustering approach for LUAD; (B) Consensus clustering matrix of two prognostic subtypes based on ten algorithms; (C) KM curves comparing the two subtypes; (D) KM curves for subtypes in the meta cohort; (E) Consistency analysis between CS and NTP in the TCGA-LUAD cohort; (F) Consistency between CS and PAM in the TCGA-LUAD cohort; (G) Consistency between NTP and PAM in the meta-LUAD cohort. LUAD: Landscape of lung adenocarcinoma; KM: Kaplan-Meier; CS: consensus subtype; TCGA: The Cancer Genome Atlas; PAM: partitioning around medoids; CIMLR: Cancer Integration via Multikernel Learning; SNF: Similarity Network Fusion; PINSPlus: Perturbation Clustering for data INtegration Plus; NEMO: Network-based Ensemble Method for Omics data; COCA: Consensus Clustering with multiple Algorithms; moCluster: multiple omics cluster; LRACluster: low-rank approximation cluster; IntNMF: integrative non-negative matrix factorization; NTP: Nearest Template Prediction.