Skip to main content
. 2020 Jul 12;105(9):e3207–e3223. doi: 10.1210/clinem/dgaa449

Figure 4.

Figure 4.

Novel immune classifications of pituitary adenomas (PAs) in the training set based on the patterns of immune cell infiltration analyzed using the unsupervised consensus clustering algorithm. (A) Cumulative distribution function (CDF) curves of the consensus score (k = 2-9). (B) Relative change in the area under the CDF curve (k = 2-9). (C) Consensus clustering matrix for k = 3, which was the optimal cluster number. (D) Principal component analysis (PCA) of the tumor-infiltrating immune cell (TIIC) patterns according to intrinsic subtyping. (E) PCA according to the novel immune clusters. (F) Distributions of the 5 intrinsic PA subtypes within the novel immune clusters. (G) Heatmap representing the distributions of 24 types of TIICs among the 3 novel immune clusters.