Fig. 4From: Classification of breast cancer patients using somatic mutation profiles and machine learning approachesConsensus matrix. The consensus matrix is 358 × 358 in size and illustrating the stability of the clustering. In ideal case, all the entries are expected to be either 0 or 1, making solid colored blocks. The bar on top indicates the clinical stage of each patient. The Silhouette score of this matrix is 0.958 which indicates a very stable clustering. (Silhouette (consensus) = 0.958)Back to article page