Table 1.
Time | Accuracy (%) | Median | Number of overlapping genes | Precision of selected genes (%) |
---|---|---|---|---|
2 | 44 | 19 | 1 | 0 |
3 | 47 | 18 | 2 | 0 |
4 | 56 | 16 | 2 | 0 |
5 | 58 | 17 | 2 | 100 |
6 | 73 | 7 | 2 | 100 |
7 | 87 | 2 | 1 | 100 |
8 | 86 | 4 | 2 | 100 |
Time is the number of time points used to construct the HMM model. Accuracy is the best accuracy using these time points (loop HMM using discriminative training). Median is the median number of selected genes. Number of overlapping genes presents the number of genes selected in at least 90% of the training–testing splits. Precision is the percent of overlapping genes that were indeed part of the 10 assigned differentially expressed genes.