Skip to main content

Table 3 Average AUC and number of logical rules according to chi-square test for Sfull and S30 databases.

From: A discriminative method for family-based protein remote homology detection that combines inductive logic programming and propositional models

Methods S full
AUC (#logical rules)
S30
AUC (#logical rules)
  no chi-square δ = 0.05 δ = 0.25 no chi-square δ = 0.5 δ = 0.25
ILP-SVM-Seq 0.79 (228.59) 0.79 (89.15) 0.75 (59.30) 0.77 (311.09) 0.77 (91.04) 0.70 (26.4)
ILP-SVM-Aln cons 0.81 (44.91) 0.81 (34.98) 0.77 (12.76) 0.81 (56.72) 0.80 (36.44) 0.72 (13.16)
ILP-SVM-Aln ps 0.80 (191.65) 0.79 (139.61) 0.75 (66.15) 0.81 (241.96) 0.81 (178.72) 0.73 (71)
ILP-SVM-Seq-Aln cons 0.85 (311.09) 0.83 (144.07) 0.79 (35.8) 0.80 (381.12) 0.80 (178.56) 0.74 (49.96)
ILP-SVM-Aln cons -Aln ps 0.82 (236.56) 0.82 (174.59) 0.79 (46.04) 0.82 (283.12) 0.82 (209.76) 0.80 (57.28)
ILP-SVM-Seq-Aln cons -Aln ps 0.87 (502.74) 0.85 (283.69) 0.81 (74.3) 0.82 (623.56) 0.82 (357.28) 0.79 (90.96)
ILP-DT-Seq 0.67 (228.59) 0.67 (89.15) 0.62 (59.30) 0.65 (311.09) 0.65 (91.04) 0.61 (26.4)
ILP-DT-Aln cons 0.70 (44.91) 0.70 (34.98) 0.72 (12.76) 0.69 (56.72) 0.69 (36.44) 0.65 (13.16)
ILP-DT-Aln ps 0.68 (191.65) 0.68 (139.61) 0.64 (66.15) 0.67 (241.96) 0.67 (178.72) 0.62 (71)
ILP-DT-Seq-Aln cons 0.72 (311.09) 0.71 (144.07) 0.67 (35.8) 0.69 (381.12) 0.68 (178.56) 0.63 (49.96)
ILP-DT-Aln cons -Aln ps 0.71 (236.56) 0.71 (174.59) 0.73 (46.04) 0.71 (283.12) 0.70 (209.76) 0.62 (57.28)
ILP-DT-Seq-Aln cons -Aln ps 0.74 (502.74) 0.74 (283.69) 0.69 (74.3) 0.71 (623.56) 0.71 (357.28) 0.63 (90.96)