Table 1.
Results from Linkage Hierarchical Clustering Analysis by Using the Each of the 10 Peak Sets We Obtained and the 7 Most Frequently Selected Peaksa
| number of samples in each cluster (percentage in the sample group) |
||||
|---|---|---|---|---|
| peak sets (number of peaks) |
sample groups |
HCC cluster |
CLD cluster |
healthy cluster |
| 447 peaks (447) | HCC | 67 (92%) | 3 (4%) | 3 (4%) |
| CLD | 4 (8%) | 24 (46%) | 24 (46%) | |
| Healthy | 27 (35%) | 14 (18%) | 37 (47%) | |
| Peak set 1 (9) | HCC | 72 (99%) | 0 (0%) | 1 (1%) |
| CLD | 0 (0%) | 36 (69%) | 16 (31%) | |
| Healthy | 12 (15%) | 16 (21%) | 50 (64%) | |
| Peak set 2 (9) | HCC | 69 (95%) | 0 (0%) | 4 (5%) |
| CLD | 0 (0%) | 32 (62%) | 20 (38%) | |
| Healthy | 9 (11%) | 14 (18%) | 55 (71%) | |
| Peak set 3 (12) | HCC | 70 (96%) | 0 (0%) | 3 (4%) |
| CLD | 0 (0%) | 25 (48%) | 27 (52%) | |
| Healthy | 10 (13%) | 10 (13%) | 58 (74%) | |
| Peak set 4 (11) | HCC | 72 (99%) | 0 (0%) | 1 (1%) |
| CLD | 0 (0%) | 37 (71%) | 15 (29%) | |
| Healthy | 9 (11%) | 20 (26%) | 49 (63%) | |
| Peak set 5 (11) | HCC | 71 (97%) | 0 (0%) | 2 (3%) |
| CLD | 0 (0%) | 32 (62%) | 20 (38%) | |
| Healthy | 12 (15%) | 17 (22%) | 49 (63%) | |
| Peak set 6 (9) | HCC | 69 (95%) | 0 (0%) | 4 (5%) |
| CLD | 0 (0%) | 32 (62%) | 20 (38%) | |
| Healthy | 9 (11%) | 14 (18%) | 55 (71%) | |
| Peak set 7 (7) | HCC | 66 (90%) | 0 (0%) | 7 (10%) |
| CLD | 0 (0%) | 48 (92%) | 4 (8%) | |
| Healthy | 7 (9%) | 21 (27%) | 50 (64%) | |
| Peak set 8 (9) | HCC | 70 (96%) | 0 (0%) | 3 (4%) |
| CLD | 0 (0%) | 26 (50%) | 26 (50%) | |
| Healthy | 8 (10%) | 8 (10%) | 62 (80%) | |
| Peak set 9 (7) | HCC | 66 (90%) | 0 (0%) | 7 (10%) |
| CLD | 0 (0%) | 48 (92%) | 4 (8%) | |
| Healthy | 7 (9%) | 21 (27%) | 50 (64%) | |
| Peak set 10 (9) | HCC | 70 (96%) | 0 (0%) | 3 (4%) |
| CLD | 0 (0%) | 47 (90%) | 5 (10%) | |
| Healthy | 13 (16%) | 10 (13%) | 55 (71%) | |
| 7 peaks selected by all 10 sets (7) | HCC | 66 (90%) | 0 (0%) | 7 (10%) |
| CLD | 0 (0%) | 48 (92%) | 4 (8%) | |
| Healthy | 7 (9%) | 21 (27%) | 50 (64%) | |
a
For each of the three clusters, the sample group with the highest number of samples is marked in boldface.