350
A. Zakharov and A. Lagunin
tion to the withdrawn drugs, seven currently marketed drugs were also analyzed to
find out the difference in the number of predicted antitargets. The prediction results
are presented in Table 11.13.
The results have shown that the number of predicted antitargets for withdrawn
drugs is considerably higher than one for marketed drugs. Thus, this service can
successfully be applied for the selection and prioritization of safe compounds during the drug discovery process.
Activity name
N
AUC, %
Lysine carboxypeptidase inhibitor
35
99.8
MAO A inhibitor
463
98.2
MAO inhibitor
1124
97.1
Na+K + transporting ATPase inhibitor
165
98.5
Neutral endopeptidase inhibitor
698
99.5
Opioid delta receptor antagonist
2044
99.0
Opioid mu receptor agonist
133
95.7
Ornithine carbamoyltransferase inhibitor
40
99.0
Peroxidase inhibitor
16
93.1
Phenylalanine 4-hydroxylase inhibitor
30
97.9
Phosphodiesterase inhibitor
5905
97.1
Phosphofructokinase-1 inhibitor
33
98.5
Phosphoglycerate kinase inhibitor
189
99.6
Phospholipase A2 inhibitor
541
96.5
Phospholipase C inhibitor
51
92.8
Phosphorylase inhibitor
279
98.5
Platelet activating factor antagonist
2431
97.2
Prostaglandin F2 alpha agonist
26
100.0
Protein kinase C stimulant
61
98.8
Pyruvate kinase inhibitor
38
97.3
Retinoic acid alpha receptor agonist
27
99.7
S-adenosyl-L-homocysteine hydrolase inhibitor
69
100.0
Sarcoplasmic reticulum calcium ATPase inhibitor
11
94.3
Sodium channel blocker
1004
93.8
Succinate dehydrogenase inhibitor
14
95.7
Superoxide dismutase inhibitor
15
88.4
Thyroid hormone agonist
104
98.6
Thyroid hormone beta agonist
74
100.0
Topoisomerase I inhibitor
478
96.3
Topoisomerase II alpha inhibitor
44
96.7
Triose-phosphate isomerase inhibitor
23
98.6
Tyrosine 3 hydroxylase inhibitor
12
88.4
UDP-glucose 4-epimerase inhibitor
28
97.2
Ubiquinol-cytochrome-c reductase inhibitor
10
81.6
N number of active compounds in the training set AUC Area Under Curve, calculated by leaveone-out cross-validation procedure
Table 11.12 (continued)
A. Zakharov and A. Lagunin
tion to the withdrawn drugs, seven currently marketed drugs were also analyzed to
find out the difference in the number of predicted antitargets. The prediction results
are presented in Table 11.13.
The results have shown that the number of predicted antitargets for withdrawn
drugs is considerably higher than one for marketed drugs. Thus, this service can
successfully be applied for the selection and prioritization of safe compounds during the drug discovery process.
Activity name
N
AUC, %
Lysine carboxypeptidase inhibitor
35
99.8
MAO A inhibitor
463
98.2
MAO inhibitor
1124
97.1
Na+K + transporting ATPase inhibitor
165
98.5
Neutral endopeptidase inhibitor
698
99.5
Opioid delta receptor antagonist
2044
99.0
Opioid mu receptor agonist
133
95.7
Ornithine carbamoyltransferase inhibitor
40
99.0
Peroxidase inhibitor
16
93.1
Phenylalanine 4-hydroxylase inhibitor
30
97.9
Phosphodiesterase inhibitor
5905
97.1
Phosphofructokinase-1 inhibitor
33
98.5
Phosphoglycerate kinase inhibitor
189
99.6
Phospholipase A2 inhibitor
541
96.5
Phospholipase C inhibitor
51
92.8
Phosphorylase inhibitor
279
98.5
Platelet activating factor antagonist
2431
97.2
Prostaglandin F2 alpha agonist
26
100.0
Protein kinase C stimulant
61
98.8
Pyruvate kinase inhibitor
38
97.3
Retinoic acid alpha receptor agonist
27
99.7
S-adenosyl-L-homocysteine hydrolase inhibitor
69
100.0
Sarcoplasmic reticulum calcium ATPase inhibitor
11
94.3
Sodium channel blocker
1004
93.8
Succinate dehydrogenase inhibitor
14
95.7
Superoxide dismutase inhibitor
15
88.4
Thyroid hormone agonist
104
98.6
Thyroid hormone beta agonist
74
100.0
Topoisomerase I inhibitor
478
96.3
Topoisomerase II alpha inhibitor
44
96.7
Triose-phosphate isomerase inhibitor
23
98.6
Tyrosine 3 hydroxylase inhibitor
12
88.4
UDP-glucose 4-epimerase inhibitor
28
97.2
Ubiquinol-cytochrome-c reductase inhibitor
10
81.6
N number of active compounds in the training set AUC Area Under Curve, calculated by leaveone-out cross-validation procedure
Table 11.12 (continued)