Table 2. Comparison of Argot2, Argot2.5, and Wei2GO raw prediction counts.

All predicted protein-GO term pairs present in the test set are classified as a True Positive (TP) and all predicted protein-GO term pairs not present in the test set are classified as False Positive (FP). Shown are the numbers for Biological Process ontology (BPO), Molecular Function ontology (MFO), and Cellular Component ontology (CCO).

	BPO TP	BPO FP	MFO TP	MFO FP	CCO TP	CCO FP
Argot2	39,781	1,766,545	3,325	122,286	19,325	238,918
Argot2.5	29,291	104,322	3,677	177,220	17,212	48,230
Wei2GO	39,460	119,702	3,467	33,815	10,903	33,748