Skip to main content

Table 1 Performance of NEMII based on different feature combinations

From: A network embedding-based multiple information integration method for the MiRNA-disease association prediction

  AUPR AUC F1 ACC REC SPEC PRE
combination 1 0.6036 ± 0.0018 0.9252 ± 0.0014 0.6072 ± 0.0020 0.9955 ± 0.0001 0.4860 ± 0.0052 0.9992 ± 0.0001 0.8128 ± 0.0158
combination 2 0.2630 ± 0.0032 0.7890 ± 0.0056 0.3338 ± 0.0032 0.9933 ± 0.0000 0.2360 ± 0.0025 0.9987 ± 0.0000 0.5681 ± 0.0058
combination 3 0.6086 ± 0.0015 0.9284 ± 0.0012 0.6129 ± 0.0031 0.9956 ± 0.0001 0.4887 ± 0.0069 0.9992 ± 0.0001 0.8247 ± 0.0160
combination 4 0.6085 ± 0.0024 0.9262 ± 0.0018 0.6115 ± 0.0026 0.9956 ± 0.0000 0.4836 ± 0.0055 0.9993 ± 0.0001 0.8366 ± 0.0105
combination 5 0.6104 ± 0.0012 0.9293 ± 0.0017 0.6147 ± 0.0025 0.9956 ± 0.0001 0.4893 ± 0.0060 0.9993 ± 0.0001 0.8289 ± 0.0164
  1. * combination 1: SDNE feature alone
  2. * combination 2: miRNA-family feature and disease similarity feature
  3. * combination 3: SDNE feature and miRNA-family feature
  4. * combination 4: SDNE feature and disease similarity feature
  5. * combination 5: SDNE feature, miRNA-family feature and disease similarity feature