Table 1. Identification of the optimal transcript subset for molecular subtype representation.
No. of top transcripts t chosen for each metagene | Transcript subseta | Adjusted Rand Index (ARI) | No. of sample misclassification |
50 | 137 | 0.94596 | 4 |
100 | 255 | 0.92589 | 5 |
150 | 366 | 0.9869 | 1 |
200 | 475 | 0.97929 | 2 |
250 | 583 | 0.97929 | 2 |
300 | 681 | 0.97929 | 2 |
400 | 846 | 0.97929 | 2 |
800 | 1116 | 1 | 0 |
Number of transcripts that are union of the top t x 3 transcripts for 3 metagenes.