A QSAR study on 2-(4-methylpiperazin-1-yl)quinoxalines as human histamine H₄ receptor ligands

Figures & data

Table 1.  Structures and observed experimental binding affinity of N-methylpiperazinylquinoxaline derivatives.^a

Cpd.	R1	R2	pKi
1^b	H	H	6.05
2	H	CH3	6.70
3	H	C6H5	4.99
4^c	H	CH2C6H5	—
5^b	H	SPh	6.44
6	H	NHCH2Ph	5.13
7	H	OPh	6.49
8	H	OCH2Ph	6.53
9	H	O(CH2)2Ph	6.32
10^b	H	O(CH2)4Ph	5.53
11	H	OCH2-2-pyridyl	6.33
12	H	OCH2-3-pyridyl	5.53
13	H	OCH2-4-pyridyl	5.36
14	H	OCH2-(4-OCH3-Ph)	6.15
15	H	OCH2-(4-CH3-Ph)	5.86
16	H	OCH2-(4-Cl-Ph)	5.64
17^b	H	OCH2-(3-Cl-Ph)	6.57
18^b	H	O(3-pyridyl)	5.89
19^b	H	O(4-Cl-Ph)	5.63
20^b	H	O(3,4-Cl-Ph)	5.77
21	H	O(4-F-Ph)	5.80
22	H	O(3-CH3-Ph)	6.24
23	H	O(4-CH3-Ph)	5.66
24	H	O-(4-OCH3-Ph)	5.63
25	H	O-(3-N,N-dimethylaniline)	5.81
26	H	O-cyclohexyl	4.88
27	H	OCH2CH(CH3)2	5.24
28	H	OCH2CH3	6.64
29	H	O(CH2)3-N,N-dimethylamine	5.40
30	H	OCH3	6.47
31	H	OH	7.21
32	6-Cl	OCH3	7.58
33^b	6-Cl	OH	7.93
34	6,7-di-Cl	OH	8.25
35^b	6,7-di-Cl	(CH2)2Ph	5.40
36	6,7-di-Cl	CH3	7.20
37^b	6,7-di-Cl	OPh	5.93
38^b	6,7-di-Cl	OCH3	7.24
39	H	CF3	5.60
40^b	H	Cl	6.64
41^b	6-Cl	H	7.04
42			5.16
43			6.23
44			4.69
45^b			5.12

^aTaken from reference [16].

^bCompounds in test set.

^cNot part of the dataset.

Table 2.  Physical meaning, average regression coefficients and the total incidences, and MLR-like coefficients from PLS model of descriptors identified from four-parameter CP-MLR models for the binding affinity of N-methylpiperazinylquinoxaline derivatives.

S. No.	Class	Descriptors’ symbol, and meaning	Avg. reg. coeff.(incidence)^a	MLR-like coeff.(fc)order^b
1	CONST	nF, number of fluorine atoms	−0.303(2)	−0.065(−0.027)10
2	TOPO	MSD, mean square distance index (Balaban)	27.145(4)	−0.031(−0.013)12
3		X2A, average connectivity index chi-2	−54.948(1)	0.004(0.002)17
4		PW4, path/walk 4-Randic shape index	57.145(2)	0.008(0.004)16
5		PW5, path/walk 5-Randic shape index	78.401(2)	0.0217(0.009)13
6		CIC5, complementary information content of fifth order neighbourhood symmetry	−2.385(1)	−0.252(−0.106)4
7	BCUT	BELm1, lowest eigen value n.1 of Burden matrix weighed by atomic masses	−30.081(19)	−0.767(−0.323)1
8	GALVEZ	GGI9, ninth order Galvez topological charge index	−4.997(13)	−0.256(−0.108)3
9		JGI2, second order mean topological charge index	17.523(2)	0.127(0.054)7
10		JGI7, seventh order mean topological charge index	−84.510(2)	−0.161(−0.068)5
11	2D-AUTO	MATS4v, Moran autocorrelation of lag 4 weighed by atomic van der Waals volumes	−3.763(2)	−0.085(−0.036)9
12		MATS6e, Moran autocorrelation of lag 6 weighed by atomic Sanderson electronegativities	1.763(1)	−0.004(−0.002)18
13		MATS7e, Moran autocorrelation of lag 7 weighed by atomic Sanderson electronegativities	2.502(2)	0.091(0.038)8
14		GATS7e, Geary autocorrelation of lag 7 weighed by atomic Sanderson electronegativities	2.324(1)	−0.015(−0.006)14
15		MATS8p, Moran autocorrelation of lag 8 weighed by atomic polarizabilities	−0.790(4)	−0.060(−0.025)11
16	ACF	C-002, CH2R2	−0.231(5)	−0.133(−0.056)6
17		C-027, R-CH-X	−0.347(8)	−0.284(−0.120)2
18		H-046, H attached to C0(SP^Citation3), no heteroatom (X) attached to next carbon	−0.127(5)	−0.011(−0.005)15

^aThe average regression coefficient of the descriptor corresponding to all models and the total number of its incidences; the arithmetic sign of the coefficient represents the actual sign of the regression coefficient in the models.

^bMLR like regression coefficient of four-component PLS model; (fc) is fraction contribution of the regression coefficient to the activity; order indicates the order of their significance in the PLS model; the constant term of PLS model is 6.024; number of compounds are 30. PLS regression and validation statistics: r = 0.955, s = 0.265, F = 65.153, Q²_LOO = 0.872, Q²_L5O = 0.883, r²_Test = 0.666.

Table 3.  Observed and modeled binding affinity of N-methylpiperazinylquinoxaline derivatives included in the training set.

Cpd.^a	pKi^b
	Obsd.	Eq. (5)		Eq. (6)		Eq. (7)		Eq. (8)		PLS
		Calc.	LOO	Calc.	LOO	Calc.	LOO	Calc.	LOO	Calc.	LOO
2	6.70	6.36	6.33	6.34	6.31	6.48	6.46	6.45	6.39	6.41	6.37
3	4.99	4.76	4.65	4.66	4.50	4.18	3.52	4.23	3.85	4.49	4.25
6	5.13	5.88	5.91	5.82	5.86	5.89	5.92	5.56	5.60	5.70	5.73
7	6.49	5.96	5.92	6.19	6.18	6.08	6.04	5.85	5.70	6.25	6.21
8	6.53	6.05	6.02	6.12	6.11	6.09	6.07	5.77	5.70	6.06	6.03
9	6.32	6.25	6.24	6.01	5.99	6.00	5.97	6.14	6.11	6.20	6.19
11	6.33	5.80	5.73	6.20	6.20	6.12	6.11	5.89	5.85	6.10	6.08
12	5.53	5.60	5.65	6.26	6.29	6.12	6.15	5.69	5.70	5.64	5.65
13	5.36	5.46	5.53	6.22	6.25	6.12	6.16	5.79	5.83	5.47	5.48
14	6.15	5.81	5.78	5.83	5.81	6.02	6.02	6.03	6.01	5.89	5.86
15	5.86	5.55	5.52	5.48	5.42	5.80	5.80	5.92	5.92	5.68	5.67
16	5.64	5.82	5.83	5.88	5.90	5.87	5.89	6.05	6.09	5.99	6.01
21	5.8	5.91	5.93	5.86	5.87	5.77	5.77	5.86	5.87	5.99	6.01
22	6.24	5.77	5.73	5.95	5.93	5.91	5.88	5.71	5.68	6.25	6.25
23	5.66	5.51	5.49	5.65	5.65	5.74	5.74	5.83	5.85	5.81	5.82
24	5.63	5.59	5.58	5.54	5.53	6.01	6.03	5.85	5.87	5.93	5.94
25	5.81	5.68	5.64	5.49	5.41	5.69	5.68	5.79	5.79	5.76	5.75
26	4.88	5.84	5.91	4.91	5.26	4.93	5.72	5.80	6.02	5.07	5.10
27	5.24	5.79	5.85	6.03	6.07	5.64	5.71	5.18	5.16	5.15	5.11
28	6.64	6.73	6.74	6.90	6.94	6.87	6.91	6.43	6.42	6.75	6.76
29	5.4	5.96	6.04	5.58	5.60	5.47	5.49	5.81	5.84	5.29	5.25
30	6.47	6.83	6.88	6.89	6.95	6.97	7.08	6.96	7.08	6.78	6.83
31	7.21	6.64	6.58	6.76	6.71	6.92	6.88	7.04	7.00	7.05	7.02
32	7.58	7.62	7.63	7.54	7.53	7.46	7.43	7.43	7.38	7.50	7.48
34	8.25	8.21	8.19	7.88	7.74	7.71	7.53	8.06	7.96	8.21	8.19
36	7.2	7.26	7.26	7.12	7.11	7.10	7.08	7.10	7.05	7.25	7.27
39	5.6	5.73	5.84	5.59	5.54	5.51	5.50	5.81	5.93	5.58	5.58
42	5.16	5.17	5.17	5.15	5.15	5.32	5.41	5.43	5.52	5.55	5.78
43	6.23	6.30	6.32	5.95	5.90	5.73	5.50	6.12	6.10	6.16	6.13
44	4.69	4.89	4.99	4.92	5.02	5.20	5.45	5.11	5.33	4.75	4.79

^aAs in Table 1.

^bOn molar basis.

Table 4.  Observed, modelled, and residual^a binding affinity of N-methylpiperazinylquinoxaline derivatives included in the test set.

Cpd.^b	pKi^c
	Obsd.	Eq. (5)		Eq. (6)		Eq. (7)		Eq. (8)		PLS
		Calc.	Resi.	Calc.	Resi.	Calc.	Resi.	Calc.	Resi.	Calc.	Resi.
1	6.05	5.96	0.09	6.55	−0.50	6.74	−0.69	6.81	−0.76	6.25	−0.20
5	6.44	5.95	0.49	6.22	0.22	6.28	0.16	5.90	0.54	6.29	0.15
10	5.53	5.85	−0.32	5.00	0.53	5.87	−0.34	6.06	−0.53	5.66	−0.13
17	6.57	5.87	0.70	5.76	0.81	6.02	0.55	5.94	0.63	6.28	0.29
18	5.89	5.34	0.55	6.26	−0.37	6.11	−0.22	5.96	−0.07	5.85	0.04
19	5.63	5.79	−0.16	5.79	−0.16	5.77	−0.14	5.93	−0.30	6.05	−0.42
20	5.77	6.03	−0.26	5.75	0.02	5.70	0.07	6.09	−0.32	6.43	−0.66
33	7.93	7.41	0.52	7.41	0.52	7.38	0.55	7.85	0.08	7.75	0.18
35	5.40	5.75	−0.35	5.01	0.39	5.21	0.19	5.59	−0.19	5.47	−0.07
37	5.93	7.03	−1.10	6.92	−0.99	6.78	−0.85	6.70	−0.77	7.01	−1.08
38	7.24	8.31	−1.07	7.99	−0.75	7.78	−0.54	7.89	−0.65	7.88	−0.64
40	6.64	6.63	0.01	6.86	−0.22	7.09	−0.45	6.91	−0.27	6.96	−0.32
41	7.04	7.12	−0.08	7.38	−0.34	7.28	−0.24	7.64	−0.60	7.11	−0.07
45	5.12	5.27	−0.15	5.87	−0.75	6.23	−1.11	5.98	−0.86	5.75	−0.63

^aDifference of observed and calculated pK_i value.

^b,cSee footnote under Table 3.

Figure 1.  Plot of observed versus calculated pK_i values for the training (Δ) and test set (○) compounds. A, B, C, and D correspond, respectively, to four-parameter models (5), (6), (7), and (8) identified through CP-MLR. E corresponds to the MLR-like PLS equation from 18 CP-MLR-identified descriptors.

Figure 2.  Plot of fraction contribution of MLR-like PLS coefficients (normalized) of the 18 descriptors (Table 2) to the activity. The 10 most significant descriptors are identified by black-shaded lines. The numerals on bars are the order of the descriptors.

Table 5.  Resultant models for the whole dataset (n = 44) in descriptors of training set models.

Model	r	s	F	Q^2LOO	Q^2L5O	Eq.
pKi = 56.188 −24.333(3.726)BELm1 −5.094(0.907)GGI9 + 1.977(0.661)MATS7e −0.310(0.118)C-027	0.860	0.437	27.774	0.675	0.675	5a
pKi = 63.082 −27.568(3.513)BELm1 −4.565(0.928)GGI9 −0.672(0.277)GATS7e −0.172(0.076)C-002	0.854	0.446	26.159	0.656	0.660	6a
pKi = 56.111 + 19.364(4.707)MSD +55.586(13.834)PW4 −31.839(3.484)BELm1 −0.217(0.074)C-002	0.854	0.445	26.339	0.646	0.582	7a
pKi = 56.984 −25.527(3.655)BELm1 +17.335(4.633)JGI2 −82.597(19.769)JGI7 +2.159(0.653)MATS8p	0.853	0.447	26.073	0.644	0.617	8a

Figure 3.  Williams plot for the training set and external prediction set for H₄R binding affinity of N-methylpiperazinylquinoxaline derivatives.

Celanire S, Wijtmans M, Talaga P, Leurs R, de Esch IJP. Histamine H3 receptor antagonists reach out for the clinic. Drug Discov Today 2005;23/24:1613–1627.

 Google Scholar