This pipeline calculates clusters based on consensus hierarchical clustering with agglomerative average linkage##REF##1,##REF##2. This pipeline has the following features:
-
Classify samples into consensus clusters.
-
Determine differentially expressed marker proteins for each subtype.
The 150 most variable proteins were selected. Consensus average linkage hierarchical clustering of 195 samples and 150 proteins identified 4 subtypes with the stability of the clustering increasing for k = 2 to k = 8 and the average silhouette width calculation for selecting the robust clusters.
Figure 1. Get High-res Image Samples were separated into 4 clusters. Shown are 195 samples and 420 marker proteins. The color bar of the row indicates the marker proteins for the corresponding cluster.
Figure 2. Get High-res Image Heatmap with a standard hierarchical clustering for 195 samples and the 150 most variable proteins.
Figure 3. Get High-res Image The consensus matrix after clustering shows 4 clusters with limited overlap between clusters.
Figure 4. Get High-res Image The correlation matrix also shows 4 clusters.
Figure 5. Get High-res Image The silhouette width was calculated for each sample and each value of k. The left panel shows the average silhouette width across all samples for each tested k (left panel). The right panels shows assignments of clusters to samples and the silhouette width of each sample for the most robust clustering.
Table 1. Get Full Table List of samples with 4 subtypes and silhouette width.
| SampleName | cluster | silhouetteValue |
|---|---|---|
| TCGA-21-1070-01A-21-2193-20 | 1 | 0.3 |
| TCGA-21-1071-01A-21-2193-20 | 1 | 0.36 |
| TCGA-21-1081-01A-21-2193-20 | 1 | 0.32 |
| TCGA-21-5782-01A-21-2194-20 | 1 | 0.041 |
| TCGA-21-5786-01A-21-2194-20 | 1 | 0.38 |
| TCGA-22-0940-01A-21-2193-20 | 1 | 0.18 |
| TCGA-22-4607-01A-21-2194-20 | 1 | 0.27 |
| TCGA-22-5471-01A-21-2194-20 | 1 | 0.096 |
| TCGA-22-5479-01A-11-2194-20 | 1 | 0.18 |
| TCGA-22-5492-01A-21-2194-20 | 1 | 0.17 |
Table 2. Get Full Table List of samples belonging to each cluster in different k clusters.
| TCGA_ID | K.2 | K.3 | K.4 | K.5 | K.6 | K.7 | K.8 |
|---|---|---|---|---|---|---|---|
| TCGA-18-4721-01A-21-2193-20 | 1 | 1 | 3 | 5 | 2 | 6 | 8 |
| TCGA-18-5592-01A-21-2194-20 | 1 | 1 | 3 | 5 | 2 | 6 | 8 |
| TCGA-21-1072-01A-21-2193-20 | 1 | 1 | 2 | 3 | 3 | 4 | 5 |
| TCGA-21-1075-01A-21-2193-20 | 1 | 1 | 2 | 3 | 3 | 4 | 5 |
| TCGA-21-1076-01A-41-2193-20 | 1 | 1 | 3 | 5 | 2 | 6 | 8 |
| TCGA-21-1078-01A-21-2193-20 | 1 | 1 | 2 | 3 | 3 | 4 | 5 |
| TCGA-21-1079-01A-21-2193-20 | 1 | 1 | 2 | 3 | 3 | 4 | 5 |
| TCGA-21-1080-01A-21-2193-20 | 1 | 1 | 3 | 5 | 2 | 6 | 8 |
| TCGA-21-1082-01A-21-2193-20 | 1 | 1 | 3 | 5 | 2 | 6 | 8 |
| TCGA-21-1083-01A-21-2193-20 | 1 | 1 | 3 | 5 | 2 | 6 | 8 |
Samples most representative of the clusters, hereby called core samples were identified based on positive silhouette width, indicating higher similarity to their own class than to any other class member. Core samples were used to select differentially expressed marker proteins for each subtype by comparing the subclass versus the other subclasses, using Student's t-test.
Table 3. Get Full Table List of marker proteins with p <= 0.05 (The positive value of column difference means protein is upregulated in this subtype and vice versa).
| PAI.1.PAL.1.M.C | X1.97649879918515e.17 | X1.68082528602647 | X1.14636930352739e.15 | X1 |
|---|---|---|---|---|
| IGFBP2|IGFBP2-R-V | 1.8e-06 | 1 | 7.6e-06 | 1 |
| PTGS2|COX-2-R-C | 0.000085 | 0.86 | 0.00024 | 1 |
| FN1|FIBRONECTIN-R-C | 1.1e-07 | 0.75 | 5.6e-07 | 1 |
| EIF4EBP1|4E-BP1-R-V | 2.8e-07 | 0.74 | 1.3e-06 | 1 |
| CHEK1|CHK1-R-C | 1.6e-14 | 0.57 | 3.5e-13 | 1 |
| CDKN1A|P21-R-C | 2.1e-12 | 0.54 | 3.3e-11 | 1 |
| MAPT|TAU-M-C | 1.9e-06 | 0.51 | 7.8e-06 | 1 |
| VASP|VASP-R-C | 1.2e-13 | 0.49 | 2.3e-12 | 1 |
| BAX|BAX-R-V | 1.8e-09 | 0.45 | 1.4e-08 | 1 |
| CD49|CD49B-M-V | 0.000021 | 0.43 | 0.000069 | 1 |
The RPPA Level 3 data was as the input file for clustering, which is Protein measurements correted by using median centering across antibodies.
Consensus Hierarchical clustering is a resampling-based clustering. It provides for a method to represent the consensus across multiple runs of a clustering algorithm and to assess the stability of the discovered clusters. To this end, perturbations of the original data are simulated by resampling techniques ##REF##1,##REF##2.
Silhouette width is defined as the ratio of average distance of each sample to samples in the same cluster to the smallest distance to samples not in the same cluster. If silhouette width is close to 1, it means that sample is well clustered. If silhouette width is close to -1, it means that sample is misclassified ##REF##3.
This is an experimental feature. The full results of the analysis summarized in this report can be downloaded from the TCGA Data Coordination Center.