Skip to main content

Table 1 Performance comparison on 5859 MultiLoc protein dataset

From: Gene ontology based transfer learning for protein subcellular localization

Subcellular location Size MultiLoc MultiLoc-GO GO-TLM
    SP SE MCC - SP SE MCC
Fungi
(5407)
cytoplasm 1411 0.68 0.85 0.70 - 0.9181 0.9773 0.9306
  endoplasmic 198 0.71 0.59 0.63 - 0.9611 0.8737 0.9136
  extracellular 843 0.73 0.81 0.73 - 0.9915 0.9656 0.9749
  golgi 150 0.71 0.53 0.60 - 0.9530 0.9467 0.9485
  mitochondria 510 0.88 0.82 0.83 - 0.9723 0.9647 0.9655
  nucleus 837 0.81 0.74 0.73 - 0.9754 0.9486 0.9558
  peroxisome 157 0.68 0.30 0.43 - 0.9735 0.9363 0.9535
  plasmamembrane 1238 0.76 0.89 0.78 - 0.9869 0.9774 0.9774
  vacuole (Astrik) 63 0.76 0.24 0.42 - 0.9355 0.9206 0.9273
Plant
(5856)
chloroplast (Astrik) 449 0.88 0.85 0.85 - 1.0000 0.9911 0.9952
  Overall Accuracy 74.60% 89.60% 96.65%
Animal
(5547)
cytoplasm 1411 0.67 0.85 0.68 - 0.9221 0.9809 0.9340
  endoplasmic 198 0.68 0.56 0.60 - 0.9667 0.8788 0.9189
  extracellular 843 0.79 0.83 0.77 - 0.9891 0.9692 0.9753
  golgi 150 0.71 0.43 0.53 - 0.9527 0.9400 0.9448
  mitochondria 510 0.88 0.82 0.83 - 0.9723 0.9627 0.9642
  nucleus 837 0.82 0.73 0.73 - 0.9826 0.9438 0.9566
  peroxisome 157 0.71 0.31 0.44 - 0.9799 0.9299 0.9533
  plasmamembrane 1238 0.73 0.90 0.76 - 0.9765 0.9750 0.9687
  lysosome (Astrik) 103 0.69 0.36 0.48 - 0.9592 0.9126 0.9344
  Overall Accuracy 74.60% 89.60% 96.27%