A mixture of feature experts approach for protein-protein interaction prediction

BMC Bioinformatics

Table 5 Global Feature Expert Importance (by MFE) in Yeast.

	MFE R50	R50 DROP	MFE AUC	AUC DROP
P	0.2310	0.0770	0.9244	0.0219
F	0.2609	0.0471	0.8821	0.0642
S	0.3191	-0.0111	0.9459	0.0004
E	0.3022	0.0058	0.9323	0.0140
Full	0.3080		0.9463

Global feature expert importance can be measured by the decrease in AUC and R50 scores when removing the expert in the MFE method. The first column lists the four feature experts. The second and fourth column list the R50 and AUC scores when applying MFE while only using the remaining three experts. The third and fifth column list the changes between these R50 and AUC scores and the full experts version. For definition of P, F, S, E experts, see details in the 'Feature' section and Table 1.

ISSN: 1471-2105