Table 2. Result of joint identification.
| Naïve approach | Naïve approach | Hierarchical approach (DEGs→PAGs) | Hierarchical approach (DEGs→PAGs) | Model-based approach | ||
|---|---|---|---|---|---|---|
| DEGs method | M1 | M2 | M1 | M2 | M3 | |
| Leptin | p-value (FDR) | 935(0) | 124(0) | 935(0) | 124(0) | 3787(15) |
| IGF.1 | p-value (FDR) | 41(0) | 67(0) | 41(0) | 67(0) | 1418(0) |
| Adiponectin | p-value (FDR) | 0(0) | 13(0) | 0(0) | 1(0) | 1027(3) |
| Insulin | p-value (FDR) | 337(0) | 129(0) | 337(0) | 129(0) | 1953(3) |
The number of significant genes determined from three joint identification methods. The numbers in each cell show the number of significant genes without multiple comparison corrections. For the purpose of comparing each approach, significance level is fixed at 5%.The numbers in the brackets show the number of significant genes detected by FDR q-value below 5%. The model-based approach provides better power than other approaches.