Skip to main content

Table 2 Prediction on archived GOA files using BMA (see Eq. (4))

From: Interspecies gene function prediction using semantic similarity

   MicroAvgF1 MacroAvgF1 1-RankLoss Fmax RAccuracy
CC H →H 0.8328 0.7203 0.8780 0.8747 0.1542
  M →H 0.8368 0.7125 0.8751 0.8750 0.1602
  M+H →H 0.8586 0.7639 0.9808 0.8787 0.2844
  D →H 0.8316 0.7197 0.8729 0.8738 0.1536
  D+H →H 0.8524 0.7236 0.9068 0.8625 0.2530
  A →H 0.8259 0.7068 0.8632 0.8532 0.1369
  A+H →H 0.8363 0.7222 0.8797 0.8773 0.1717
  M →M 0.7712 0.6084 0.8588 0.8571 0.1936
  H →M 0.7676 0.6192 0.8580 0.8339 0.1864
  H+M →M 0.8161 0.6590 0.9548 0.8637 0.3518
  D →M 0.7718 0.6105 0.8416 0.8082 0.1868
  D+M →M 0.8003 0.6160 0.8926 0.8315 0.2963
  A →M 0.7660 0.6341 0.8444 0.8404 0.1761
  A+M →M 0.7713 0.6143 0.8606 0.8523 0.1942
MF H →H 0.8523 0.8179 0.9192 0.8915 0.1416
  M →H 0.8513 0.8170 0.9145 0.8905 0.1311
  M+H →H 0.8692 0.8372 0.9720 0.9029 0.2399
  D →H 0.8502 0.8174 0.9123 0.8909 0.1295
  D+H →H 0.8668 0.8355 0.9523 0.8742 0.2259
  A →H 0.8416 0.8207 0.8964 0.8968 0.0793
  A+H →H 0.8490 0.8151 0.9116 0.8894 0.1227
  M →M 0.7654 0.6849 0.8755 0.8656 0.1344
  H →M 0.7601 0.6821 0.8797 0.8545 0.1396
  H+M →M 0.7784 0.7081 0.9248 0.8779 0.1801
  D →M 0.7607 0.6891 0.8592 0.8369 0.1287
  D+M →M 0.7841 0.7072 0.9200 0.8580 0.2013
  A →M 0.7534 0.6880 0.8553 0.8607 0.0876
  A+M →M 0.7639 0.6716 0.8712 0.8551 0.1264
BP H →H 0.8373 0.7979 0.9507 0.8012 0.2044
  M →H 0.8346 0.7943 0.9489 0.7981 0.1912
  M+H →H 0.8450 0.8055 0.9690 0.8381 0.2421
  D →H 0.8368 0.8027 0.9568 0.8031 0.2020
  D+H →H 0.8368 0.7978 0.9496 0.8093 0.2018
  A →H 0.8290 0.7903 0.9239 0.7799 0.1641
  A+H →H 0.8325 0.7839 0.9308 0.7944 0.1809
  M →M 0.7812 0.6965 0.9350 0.7905 0.1855
  H →M 0.7842 0.6987 0.9401 0.7863 0.1965
  H+M →M 0.7947 0.7134 0.9609 0.8270 0.2357
  D →M 0.7816 0.7036 0.9348 0.7885 0.1867
  D+M →M 0.7830 0.7108 0.9423 0.7875 0.1923
  A →M 0.7768 0.6929 0.9027 0.7594 0.1692
  A+M →M 0.7779 0.6871 0.9183 0.7807 0.1733
  1. H →H directly uses GO annotations of Human to predict annotations of Human genes. M →H only employs annotations of genes from Mouse to predict annotations of Human genes. M+H →H uses GO annotations of genes from Mouse and Human to predict annotations of Human genes. D+H →H uses annotations of genes from Danio rerio and Human to predict annotations of Human genes. A+H →H uses annotations of genes from Arabidopsis thaliana and Human to predict annotations of Human genes. M →M, H+M →M, D+M →M and A+M →M follow the similar protocol, but predict annotations of Mouse genes. The data in boldface is the statistically significant best among these comparing methods for a particular target species, and the significance is checked by paired t-test at 95% confidence intervals