Table 5 Most difficult biological process and molecular function classes
From: Evaluating a variety of text-mined features for automatic protein function prediction with GOstruct
Original co-mentions | ||||||
|---|---|---|---|---|---|---|
GO ID | Name | # Predictions | Precision | Recall | F-measure | IC |
GO:0051179 | localization | 28 | 0.107 | 0.054 | 0.072 | 5.70 |
GO:0016247 | channel regulator activity | 115 | 0.043 | 0.208 | 0.071 | 6.53 |
GO:0009055 | electron carrier activity | 108 | 0.03 | 0.111 | 0.055 | 6.94 |
GO:0007067 | mitosis | 23 | 0.043 | 0.031 | 0.036 | 7.54 |
GO:0042056 | chemoattractant activity | 53 | 0.018 | 0.067 | 0.029 | 7.56 |
Enhanced co-mentions | ||||||
GO ID | Name | # Predictions | Precision | Recall | F-measure | IC |
GO:0009055 | electron carrier activity | 102 | 0.090 | 0.138 | 0.109 | 6.94 |
GO:0051179 | localization | 42 | 0.071 | 0.055 | 0.061 | 5.70 |
GO:0019838 | growth factor binding | 44 | 0.021 | 0.035 | 0.027 | 5.99 |
GO:0070888 | E-box binding | 99 | 0.010 | 0.066 | 0.019 | 7.49 |
GO:0030545 | receptor regulator activity | 152 | 0.007 | 0.020 | 0.010 | 7.63 |
Bag-of-words | ||||||
GO ID | Name | # Predictions | Precision | Recall | F-measure | IC |
GO:0051179 | localization | 18 | 0.277 | 0.090 | 0.137 | 5.70 |
GO:0009055 | electron carrier activity | 29 | 0.103 | 0.083 | 0.092 | 6.94 |
GO:0016042 | lipid catabolic process | 26 | 0.076 | 0.054 | 0.063 | 5.80 |
GO:0015992 | proton transport | 15 | 0.066 | 0.047 | 0.055 | 7.29 |
GO:0005516 | calmodulin binding | 14 | 0.071 | 0.033 | 0.045 | 7.25 |
Co-mentions + Bag-of-words | ||||||
GO ID | Name | # Predictions | Precision | Recall | F-measure | IC |
GO:0051179 | localization | 61 | 0.100 | 0.109 | 0.104 | 5.70 |
GO:0009055 | electron carrier activity | 62 | 0.079 | 0.138 | 0.101 | 6.94 |
GO:0030545 | receptor regulator activity | 63 | 0.064 | 0.080 | 0.071 | 7.63 |
GO:0042056 | chemoattractant activity | 24 | 0.041 | 0.066 | 0.051 | 7.56 |
GO:0040007 | growth | 27 | 0.030 | 0.066 | 0.047 | 7.33 |