BMC Bioinformatics

Table 4 Comparison with PAIRpred, PPiPP and multi-layered LSTM

From: Attention mechanism enhanced LSTM with residual architecture and its application for protein-protein interaction residue pairs prediction

Data set	Method		RFTP(p)
			10%	25%	50%	75%	90%
DBD 3.0	PPiPP		9	19	78	297	760
	PAIRPred
	PAIRPred _1	No post-processing	2	13	68	257	804
	PAIRPred _2	No post-processing	1	5	22	89	282
		With post-processing	1	3	16	103	272
DBD 4.0	PAIRPred _2	No post-processing	2	6	19	75	340
		With post-processing	1	3	18	101	282
DBD 5.0	Multi-layered LSTM Network	lstm _1_nodes_20	12	53	139	175	331
		lstm _5_nodes_20	13	17	46	146	271
		lstm _6_nodes_35	1	2	7	639	1384
		lstm _5_nodes_45	4	13	36	94	847
	our model	layer _60_unit_8	4	31	33	113	194

Note: lstm _m_nodes_n means the model has m layer LSTMs,and each layer has n units

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com