Experiments of 5-fold Cross-Validation on five kingdoms

*For ontology and other abbreviations, see here

Table 1: Confusion matrix for Animal group

Prd ⇒
Obs ⇓		 nuclear  mitochondrial  cytoplasmic  extracellular  golgi  peroxisomal  endoplasmic reticulum  lysosomal  membrane  No Prediction  rowSUM  recall 
nuclear 2575 2 100
 11 2
 0 3
 0 28 125
 2846 .905
mitochondrial 2 1162 18 4
 0 3 2
 0 2
 5
 1198 .970
cytoplasmic 47
 16 1695 29
 18 5 4
 2 12 17
 1845 .919
extracellular 2 0 23 3655
 1 0 4 15 150 93
 3943
 .927
golgi 1 0 5 2
 149 0 4 1 4
 1
 167 .892
peroxisomal 0 2 0 0 0 100 0 0 1 0
 103 .971
endoplasmic reticulum 0 1 4
 1
 10 0 435 0 4 2
 457 .952
lysosomal 0 0
 4 0
 0 0 0 161 4 1
 170 .947
membrane 2 11 109 59
 26 2 49
 8
 4519 35
 4820 .938
colSUM 2629 1194 1958 3761
 206 110 501 187 4724 279
 15549 Overall Recall = .929
precision .979 .973 .866 .972 .723 .909 .868 .861 .957 Sequence Coverage = .982
   Overall Precision = .946
specificity
 .996
 .998
 .981
 .991
 .996
 .999
 .996
 .998
 .981

Overall Specificity= .988

Table 2: Confusion matrix for Plant group

Prd ⇒
Obs ⇓ 		nuclear  mitochondrial  cytoplasmic  extracellular  golgi  mitochondrial  peroxisomal  endoplasmic reticulum  vacuolar  membrane  No Prediction  rowSUM   recall 
nuclear 162 0 1
 0
 1 0 0 1 0 0 3
 168 .964
mitochondrial 0
 287 6
 1 0 10 1 0 0 0 2
 307 .935
cytoplasmic 1 4 429 2 0 8
 0 0 1 0 2
 447 .960
extracellular 0 0 2
 110 0 0 0 0 3
 0
 12
 127 .866
golgi 0 0 0
 0 34 0 0 0 0 0 1
 35 .971
mitochondrial 1 18 24 1
 0
 1821 1 2 1
 18
 12
 1899 .959
peroxisomal 0 0
 0 0 0 1
 28 0 0 0 0
 29 .966
endoplasmic reticulum 0 0 1
 0 0 0 0 56 3 3 1
 64 .875
vacuolar 0 0 0
 8
 2 0 0 1
 67 3
 1
 82 .817
membrane 0 1
 2 2
 3 18
 0 2
 2
 99
 6
 135 .733
colSUM 164 310
 465 124 40 1858 30 62 77 123 40
 3293 Overall Recall = .939
precision .988 .926 .923 .887 .850 .980 .933 .903 .870 .805 Sequence Coverage = .988
   Overall Precision =  .951
specificity
 .999
 .992
 .987
 .996
 .998
 .973
 .999
 .998
 .997
 .992

Overall Specificity = .982

Table: Confusion matrix for Fungi Group
 
Prd ⇒
Obs ⇓ 		nuclear  mitochondrial  cytoplasmic  extracellular  golgi  peroxisomal  endoplasmic reticulum  membrane  vacuolar  No Prediction  rowSUM   recall 
nuclear 517 4 28
 2 1 1 0
 2 0 66
 621 .833
mitochondrial 8 302 39
 2
 0 6
 1 5 0 43
 406 .744
cytoplasmic 23 24 319 3 3 6
 3 2
 3
 9
 395 .808
extracellular 0 3
 4
 149 1 1
 1 2 2
 8
 171 .871
golgi 0
 0 3
 0 42 0 2
 1 0
 4
 52 .808
peroxisomal 0
 6 2
 0 0 55 0 0 0 1
 64 .859
endoplasmic reticulum 2 0 2
 1 6
 0 42 7 0 4
 64 .656
membrane 4
 1 7 4 6
 1 6
 260 3
 10
 302 .861
vacuolar 0 0 2 2 2 0 1 0 12 0
 19 .632
colSUM 554 340 406
 163 61 70 56 279 20
 145
 2094  Overall Recall = .811
precision .933 .888 .786 .914 .689 .786 .750 .932 .600 Sequence Coverage= .931
   Overall Precision = .871
specificity
 .975
 .977
 .949
 .993
 .991
 .993
 .993
 .989
 .996

Overall Specificity = .975

Table: Confusion matrix for Gram-positive Bacteria group

Prd ⇒
Obs ⇓ 		cytoplasmic  cell wall  extracellular  membrane  No Prediction  rowSUM  recall 
cytoplasmic 882 1
 25
 13 9
 930 .948
cell wall
 1
 15 0
 1
 2
 19
 .789
extracellular 2
 2
 222 8
 18
 252 .881
membrane 8 2
 17
 290 23
 340 .853
colSUM 893
 20 264
 312 52
 1541  Overall Recall = .914
precision .988 .750 .841 .929 Sequence Coverage = .966
   Overall Precision = .946
specificity
 .982
 .997
 .967
 .982

Overall Specificity = .980

Table: Confusion matrix for Gram-negative Bacteria group
 
Prd ⇒
Obs ⇓ 		cytoplasmic  extracellular  periplasmic  inner membrane  cell wall  outer membrane  No Prediction  rowSUM   recall 
cytoplasmic 1778 21
 26 12 0 4 20
 1861 .955
extracellular 3 217 6
 0 2 1
 24
 253 .858
periplasmic 6 15
 336 4 0 6 18
 385
 .873
inner membrane 5 1 4 411 0 0 11
 432 .951
cell wall
 0 1
 0 0 43 1 1
 46 .935
outer membrane 0 4
 2 2 0 181 8
 197 .919
colSUM 1792 259
 374 429 45 193 82
 3174  Overall Recall = .934
precision .992 .838 .898 .958 .956 .938 Sequence Coverage = .974
   Overall Precision = .959
specificity
 .989
 .986
 .986
 .993
 .999
 .996

Overall Specificity = .990

Table: Confusion matrix for Archea group

Prd ⇒
Obs ⇓ 		cytoplasmic  membrane  cell wall  extracellular No Prediction  rowSUM   recall 
cytoplasmic 401 1
 0 1 1
 404 .993
membrane 1 60 6
 0 1
 62 .968
cell wall 0 0 6 0 0 6 1.000
extracellular 0 0 0 2 3 6 .400
colSUM 402 61 6 3 5 478  Overall Recall = .983
precision .998 .984 1.000 0.667 Sequence Coverage = .990
   Overall Precision = .994
specificity
 .986 .998 1.000 .998

Overall Specificity = .988