Effect of the lipophilic parameter (log P) on the anti-parasitic activity of imidazo[1,2-a]pyridine derivatives

Article abstract of DOI:10.1007/s00044-010-9547-3

A number of imidazo[1,2-a]pyridine derivatives were selected and investigated in relation to antiparasitic (Trichomonas vaginalis) activity. After treatment with derivatives, biological activity was assessed by determination of the in vitro viability of c

Full text of DOI:10.1007/s00044-010-9547-3

MEDICINAL
CHEMISTRY
RESEARCH
Med Chem Res (2012) 21:415–420
DOI 10.1007/s00044-010-9547-3
ORIGINAL RESEARCH
Effect of the lipophilic parameter (log P) on the anti-parasitic
activity of imidazo[1,2-a]pyridine derivatives
•
Margarita Lopez-Martınez Hector Salgado-Zamora
•
´
Ma. Elena Campos-Aldrete Jose G. Trujillo-Ferrara
´
´
•
•
´
Jose Correa-Basurto Carlos Mexica-Ochoa
•
´
Received: 2 July 2010 / Accepted: 15 December 2010 / Published online: 7 January 2011
Ó Springer Science+Business Media, LLC 2010
Abstract A number of imidazo[1,2-a]pyridine deriva-
tives were selected and investigated in relation to anti-
parasitic (Trichomonas vaginalis) activity. After treatment
with derivatives, biological activity was assessed by
determination of the in vitro viability of cell cultures, using
alamar blue as a metabolic indicator. A good correlation
was found between the anti-parasitic activity and the par-
tition coefficient log P determined experimentally on the
tested compounds, which explained up to 84% of the
measured activity. A favorable interval (0.9 0.3 log
P) was found for optimum biological response.
considerable efforts are ongoing to find alternative drugs
for the treatment of parasitic diseases.
The fused heterocyclic system imidazo[1,2-a]pyridine
is recognized as an important pharmacophore, which has
been investigated in connection to a great number of
pharmacological responses. The structural resemblance of
imidazo[1,2-a]pyridine with the benzimidazole nucleus is
one of the reasons why derivatives of the former were
tested as antimicrobial (Teulade et al., 1978) as well as
anti-parasitic (Trichomonas fetus) (Winkelmann et al.,
1977) agents. On the other hand, the hydrophobic char-
acter of a molecule often seems to be a very important
physicochemical parameter for explaining the variations
Keywords Anti-parasitic activity ꢀ Log P ꢀ
Imidazo[1,2-a]pyridine derivatives
ˇ
of biological activity (Golovanov et al., 2002; Acanski
´
´
and Ðakovic-Sekulic, 2004; Eros et al., 2002; Gens-
mantel, 1994). However, a correlation of the hydrophobic
character of imidazo[1,2-a]pyridines with anti-microbial
or -parasitic activity has not been investigated. Therefore,
an investigation was carried out to explore the influence
of the lipophilic factor of a number of selected imi-
dazo[1,2-a]pyridine derivatives on anti-parasitic activity
in relation to a highly pathogenic strain (Trichomonas
vaginalis).
Introduction
Parasitic infections exist worldwide and affect large pop-
ulations, especially in underdeveloped countries where
poor hygiene conditions prevail. The appearance of anti-
parasitic drug resistant strains, in spite of prophylactic
policies, creates a more complex scenario. Therefore,
Materials and methods
´
´
M. Lopez-Martınez ꢀ H. Salgado-Zamora (&) ꢀ
Ma. E. Campos-Aldrete ꢀ C. Mexica-Ochoa
Chemistry
´
Departamento Quımica Organica, Escuela Nacional Ciencias
Biologicas, IPN, Prolongacion Carpio y Plan de Ayala S/N,
´
´
´
Most of the chemical compounds selected for this inves-
tigation are known. Some of them are commercial prod-
ucts, others have been previously described in the
literature, and a few are new compounds. Products 3–12
were prepared in house according to synthetic Scheme 1,
by applying well-established protocols. Compounds were
11340 Mexico, DF, Mexico
e-mail: hsalgado47@hotmail.com
J. G. Trujillo-Ferrara ꢀ J. Correa-Basurto
´
Laboratorio de Modelado Molecular y Bioinformatica, SEPI,
Escuela Superior de Medicina, IPN, Plan de San Luis y Dıaz
´
´
Miron, 11340 Mexico, DF, Mexico
123

416
Med Chem Res (2012) 21:415–420
3-Nitroimidazo[1,2-a]pyridine (6)
identified through their physical constants as well as
spectroscopic data. 1H and 13C NMR spectra were recorded
at 300 and 75 MHz, respectively, using a Varian Mercury
Vx 300 MHz spectrometer. New products were further
characterized by combustion analysis.
The title compound was isolated as a light yellow solid,
yield 86%, mp: 210–211 and 203–204°C (Winkelmann et
al., 1977).
3-Nitroimidazo[1,2-a]pyridine-2-carboxamide (7)
Ethyl 3-nitro imidazo[1,2-a]pyridine-2-carboxylate (3)
The title compound was prepared following a procedure in
the literature and isolated as colorless needles in 70% yield,
mp: 274–276, 265°C (Teulade et al., 1978) and 275–276°C
(Rizo et al., 2004).
The title compound was prepared following the protocol
described by Lombardino (1965) and isolated as greenish
needles, 64.7%, mp: 105–107 and 105°C (Teulade et al.,
1978).
5-Methyl-3-nitroimidazo[1,2-a]pyridine-2-carboxamide
(8)
Ethyl 5-methyl-3-nitro imidazo[1,2-a]pyridine-2-
carboxylate (4)
Following the same protocol (Rizo et al., 2004), the title
compound was obtained as mustard colored crystals, in
80% yield, mp: 212–214°C. IR mcm-1 1683, 1528. 1H
NMR (DMSOd6) d 7.76 (d, J8,7 9.2 Hz, 1H), 7.66 (dd, J7,6
7.0, J7,8 7.0 1H), 7.23 (d, J6,7 7.0, 1H), 2.5 (s, 3H), 7.6 (bs,
1H), 8.2 (bs, 1H). 13C NMR d 19.0, 114.8, 115.85, 117.75,
130.47, 137.51, 137.56, 143.35, 162.39. C9H8N4O3. Anal.
calcd. for: C, 49.09; H, 3.66; N, 25.45; found: C, 49.31; H,
3.83; N, 25.14.
The title compound was obtained following the same
protocol (Lombardino, 1965) as yellow crystals, yield 55%,
mp: 96–98 and 101°C (Winkelmann et al., 1977).
3-Nitro imidazo[1,2-a]pyridine-2-carboxylic acid (5)
The title compound was isolated as whitish crystals, yield
83%, mp: 208–209 and 206°C (Winkelmann et al., 1977).
Scheme 1 Synthetic route used
to obtain imidazo[1,2-a]
pyridine derivatives
H
Br
1.BrCH2COCO2Et/
DME/rt
2. EtOH/heat
N
CO2Et
N
N
1. R = H
R
N
NH2
2. R = Me
R
R
HNO3/H2SO4/
0-5 °C
N
N
1. aq. KOH
CO2H
CO2Et
N
2. conc. HCl
3. R = H
4. R = Me
NO2
5. R = H
R
NO2
40% H2SO4
N
reflux
NH4OH/rt/8 h
N
N
NO2
R
CONH2
N
6. R = H
7. R = H
8. R = Me
NO2
R
POCl3/110 °C/1 h
N
N
NHR1
CN Cyclohexylamine
N
N
heat
NO2
R
NO2
R
11. R = H, R1 = cyclohexyl
12. R = Me, R1 = cyclohexyl
9. R = H
10. R = Me
123

Med Chem Res (2012) 21:415–420
417
3-Nitroimidazo[1,2-a]pyridine-2-carbonitrile (9)
vitamin mix (in vitro Laboratories, Mexico) were tested on
the coupling and proliferation of the culture until a steady
growth was reached. Benzyl penicillium (2%) was incor-
porated into the medium to ensure axenic conditions.
The title compound was prepared following a procedure in
the literature (Arias et al., 2006) and isolated as reddish
crystals in 80% yield, mp: 167–168 and 162°C (Winkel-
mann et al., 1977).
Experimental conditions
5-Methyl-3-nitroimidazo[1,2-a]pyridine-2-carbonitrile
(10)
Treatment was carried out at 24 and 48 h. Controls used
were metronidazole (Rhone-Poulenk, Mexico) and a series
of seven untreated populations (400,000–150,000 tropho-
zoites), and fresh culture broth was included. In order to
show that a maximum solvent concentration in the drug
sensitivity test had no effect on the viability of the para-
sites, dimethyl sulphoxide (DMSO, biological grade,
Sigma-Aldrich, \0.03%, v/v) was included.
The title compound was obtained from 2-carboxamido-
5-methyl-3-nitroimidazo[1,2-a]pyridine, following the same
procedure (Arias et al., 2006) and isolated as light brown
crystals in 90% yield, mp: 144–146°C. IR m cm -1 2250. 1H
NMR (DMSOd6) d 7.9 (m, 2H), 7.45 (d, J6,7 6.05, 1H)
2.7(s, 3H). 13C NMR d 21.15, 112.78, 116.42, 119.01,
120.77, 121.4, 133.36, 140.03. 146.99. C9H6N4O2. Anal.
calcd. for: C, 53.47; H, 2.99; N, 27.71; found: C, 53.79; H,
3.32; N, 27.58.
Procedure
The working solutions were prepared by dissolving 10 mg
of each compound or of metronidazole, in DMSO and
diluting with the culture medium (TYI-S-33), while the
multipanel plate (96 wells bottom plane), was prepared by
addition of 100 ll of the culture medium to each well. The
operation was successively repeated so that the serial
twofold dilutions allowed seven concentrations ranging
from 0.5 to 32 lg/ml throughout the plate. Then, a sus-
pension (100 ll) of trophozoites (50,000) was incorporated
into each well and incubated during 24 h at 37°C. Controls
without drug application were included in each plate and
contained different amounts of parasites, ranging from 100
(50, 000 parasites) to 1.5% (750 parasites). Alamar blue
(20 ll) was added to each well and the plate was monitored
during 3 h.
2-(N-cyclohexyl)amino-3-nitroimidazo[1,2-a]pyridine (11)
The title compound was prepared according to the same
procedure and isolated as a yellow solid in 62% yield, mp:
156–158 and 157–159°C (Arias et al., 2006).
2-(N-cyclohexyl)amino-5-methyl-3-nitroimidazo[1,2-
a]pyridine (12)
5-Methyl-3-nitroimidazo[1,2-a]pyridine-2-carbonitrile was
heated with an excess of cyclohexylamine, following the
same procedure (Arias et al., 2006) at 60–70°C to give the
title compound as dark brown crystals in 70% yield, mp:
1
122–124°C. IR (KBr) m cm-1 3369, 1601, 1475. H NMR
(DMSOd6) d 7.54 (dd, J7,8 8.74, J7,6 7.23, 1H), 7.44 (m, 1H),
7.27 (d, J8,7 8.74, 1H), 6.8 (d, J6,7 6.46, 1H), 2.6 (s, 3H),
1.2–2.2 (m, 11H). 13CNMR d 22.5, 24.6, 25.4, 33.2, 51.3,
112.1, 115.5, 134.5, 143.0, 150.6, 155.9. C14H18N4O2.
Anal. calcd. for: C, 61.30; H, 6.61; N, 20.42; found: C,
61.01; H, 6.69; N, 20.1.
Colorimetric evaluation of trichomonicidal activity was
carried out based on the properties of the metabolic indi-
cator alamar blue. The evident end point (blue to pink) and
the intermediate color tones of the cultures allowed for
assessment of the minimal inhibitory concentration (MIC)
required to inhibit cell growth (Yajko et al., 1995). Tests
were carried out in triplicate with independent runs, under
a randomly double blind designed bioassay.
Anti-parasitic activity: in vitro evaluation
Spectrophotometric (UVWINLAB Perkin Elmer) read-
ings at k = 630 for the oxidized form (blue) of alamar blue
and at k = 570 for the reduced form (pink) were obtained
with the aid of the Beckman coulter lector software, thus
enabling the calculation of viable and non-viable tropho-
zoites in each culture. Then, the absorbance data obtained
from the micro culture readings were interpolated in the
corresponding trophozoite population calibration curve to
calculate the percentage of dead parasites. This latter value
was plotted against the concentration of the respective
compound to obtain the concentration-activity profile. The
Strain treatment
The anti-parasitic activity of the imidazo[1,2-a]pyridine
derivatives 3–12 was evaluated using Trichomonas vagi-
nalis GT3 (GT3 is a highly pathogenic strain isolated in the
City of Guanajuato, Mexico) micro-cultures in multipanel
plates. The culture medium (TYIS-33) was prepared as
described by Diamond (1995) and used for both prolifer-
ation and biological assays. Previously, fetal bovine serum
(FBS, lot A 7502 Microlab, Mexico) and the Diamond
123

418
Med Chem Res (2012) 21:415–420
compound concentration required to cause the death of
50% (EC50) of the microorganisms was estimated as well.
Table 1 Trichomonicidal activity of nitro imidazo[1,2-a]pyridine
derivatives
Trichomonicidal activity
8
1
7
6
N
Log P calculation
EC50 (lM)
MIC (lM)
R1
N
4
The partition coefficient log P was experimentally deter-
mined for each compound in a 1-octanol/water (1-octanol,
Sigma-Aldrich Analytical grade) system by applying the
shake flask methodology (Sangster, 1997). The maximum
absorbance (kmax) of each compound, metronidazole
included, was obtained by UV spectroscopy. Then, a con-
veniently stable concentration (0.85 mg/ml) of compound/
1-octanol (pre-saturated with water) was prepared as a
stock solution. From this, series of dilutions in ethyl
alcohol were obtained and read on the spectrophotometer.
The concentration of each compound was adjusted through
absorbance sweeps so that the kmax values fell within a
range of 0.2–0.8 absorbance units (concentrations fell in
the interval 0.001–0.01 mg/ml).
NO2
R2
Compounds
R1
R2
48 h
24 h
48 h
3
COOEt
COOH
H
H
3.96
21.10
2.45
2.13
19.30
3.01
2.13
19.30
3.07
5
H
6
H
7
CONH2
CN
H
41.30
9.68
9.71
9.71
9
H
5.32
2.66
11
4
NHC6H10
COOEt
CONH2
CN
H
61.53
9.11
61.50
4.02
30.80
4.02
CH3
CH3
CH3
CH3
8
14.63
12.12
64.27
1.20
9.09
9.09
10
12
4.95
2.48
NHC6H10
58.40
2.92
29.20
2.92
Triplicate test vessels containing the required, accu-
rately measured amounts of the two solvents together
[1-octanol/buffer solution (0.25 M, pH 7.4, JT Baker)]
with the necessary quantity of the stock solution were
prepared for each of the test conditions. In addition, a tube
containing an ethyl alcohol/buffer solution (blank) and a
tube containing only a solution of 1-octanol/compound
(concentrated solution) were included.
Metronidazole
Mean values of log P are shown in Table 2. Standard
deviations from the average readings of the ratios of three
volumes were not statistically significant and therefore the
method was considered reliable.
Correlation of log P and anti-parasitic activity
The extraction systems of a two-phase 1-octanol-buffer
solution were prepared in a calculated volume ratio of 1:1,
1:2 and 2:1 and then shaken for 10 min at 25 1°C in a
shaker lab-Line 3D rotator, and allowed to stand for 48 h.
Once the partition equilibrium was reached between pha-
ses, the concentration of each compound in the organic
phase was evaluated by taking an aliquot from each tube
for each test condition, and then further analyzed by UV
spectroscopy. The total quantity of substance present in
both phases was calculated by comparison with the quan-
tity of the substance originally introduced (concentrated
solution). Log P was calculated as log (ratio of the con-
centration in the 1-octanol phase to the concentration in the
aqueous phase). Metronidazole was included for compari-
son (Leo et al., 1971).
Average experimental values of log P were used to
establish the correlation with anti-parasitic activity data.
Different mathematical descriptors were tried to establish
such correlation. It was found that the mathematical
expression shown in Eq. 1 is statistically significant with
Table 2 Log P values of imidazo[1,2-a]pyridine derivatives
Compounds
Log P
1-Octanol/water volume
ratio
1:1
1:2
2:1
Log P average kmax
3
0.983
1.045
0.963
0.997 0.035 265.30
5
-0.041 -0.042 -0.038 -0.040 0.002 252.50
0.938 0.960 0.922 0.940 0.016 252.45
-0.515 -0.569 -0.540 -0.541 0.022 258.24
6
Results and discussion
7
9
0.691
1.541
1.268
0.670
1.193
1.556
0.641
1.500
1.204
0.690
1.125
1.500
0.110
0.692
1.503
1.303
0.695
1.083
1.527
0.115
0.675 0.024 257.80
1.514 0.019 230.22
1.258 0.041 269.80
0.690 0.011 265.50
1.135 0.045 263.10
1.528 0.023 235.19
0.119 0.004 310.25
The trichomonicidal activity profile was obtained as a
function of compound concentration, and the EC50 values
were determined. Molar MICs were calculated from the
minimum concentrations that inhibited proliferation of the
culture, as indicated by alamar blue. Table 1 shows the
trichomonicidal activity exhibited by the tested compounds
in terms of both parameters.
11
4
8
10
12
Metronidazole 0.120
123

Med Chem Res (2012) 21:415–420
419
exhibited anti-parasitic activity similar to that shown by
metronidazole itself.
As anticipated, the introduction of the methyl group in
the pyridine moiety increased the lipophilic character of the
imidazo[1,2-a]pyridine derivatives, evidenced by greater
log P values and higher hydrophobicity (Table 2). Log P of
metronidazole was 0.11, while compound 12 was the most
lipophilic (15.28) and the least active.
With regard to the quantitative anti-parasitic activity and
its correlation with the partition coefficient, the following
observations were made. Nitro-amide 7 was the most
hydrophilic of the series (-0.541), with a rather poor
activity. In general, the presence of the methyl group
(compounds 4, 10 and 12) diminished the anti-parasitic
activity. Remarkably, with derivative 8, in which the nitro-
carboxamide carried a methyl group, both hydrophobicity
as well as the anti-parasitic activity was increased com-
pared to the non-methylated analog 7. The most active
compounds, 3 and 6, had lipophilic values of 0.997 and
0.94, respectively. In accordance with these results, an
optimum range of lipophilic character must exist for the
compounds to show an efficient biological response.
If the partition coefficient (log P) were the determining
physical property influencing the in vitro activity of these
compounds, a bell-shaped distribution curve would have
been obtained, and all the points in Fig. 1 would fit into one
plot, indicating a standard biological effect (Hansch et al.,
1968). However, the fact that a plateau is present in the
curve shape clearly indicates that additional factors (pri-
marily electronic) should be operating as well (Butler et al.,
1967).
Fig. 1 Correlation of anti-parasitic activity with the hydrophobic
parameter log P. The STATISTICA program was used to obtain the
correlation
each of its coefficients other than zero, and therefore was
chosen as a model. The correlation coefficient, r = 0.916,
was obtained. A polynomial expression of order 4 was
examined as well. However, its coefficients were not sta-
tistically significant and therefore it was ruled out.
4
log 1=EC50 ¼ ꢁ1:29ðꢂ0:24Þ ðlogPÞ þ1:80 ðꢂ0:37Þ
3
ꢃ ðlogPÞ ꢁ1:26ðꢂ0:11Þ
ð1Þ
n ¼ 10
À
Á
ðr ¼ 0:9164Þ; r2 ¼ 0:8397 ; s ¼ 0:2159; Fð2:7Þ
¼ 18:34; P\0:0017:
Equation 1 shows that the two variables describe up to
84% of the total biological activity in accordance with the
determination of the coefficient, r2 = 0.8397. Compound 6
may be excluded from the correlation plot on the basis that
it falls 95% of the confidence limits (Moridani et al., 2004).
Furthermore, if eliminated a better coefficient would result
(an increase [0.05 units is actually obtained). However, it
was included.
Conclusion
The results indicate that ethyl 3-nitroimidazo[1,2-a]pyri-
dine-2-carboxylate 3 and the 3-nitroimidazo[1,2-a]pyridine
6 had good anti-parasitic activity, requiring 2.45–3.96 lM
to reach the desired effect.
An experimental partition coefficient (log P) determi-
nation for all the tested compounds indicated an increase in
the hydrophobicity of the methyl-substituted imidazo[1,2-
a]pyridine derivatives. A mathematical descriptor corre-
lating the exhibited pharmacological activity and log P was
found (0.9 0.3 log P), which suggests that an optimum
balance between hydrophilic and lipophilic properties is
most convenient. Calculations based on Eq. 1 show that
84% of the total activity variation found may be described
by these two variables.
Figure 1 shows the graph obtained by applying the
mathematical model described in Eq. 1. In agreement with
these results, an optimum interval of lipophilic character
does exist where the tested compounds show good bio-
logical response.
Most of the compounds showed anti-parasitic activity
within 24 h of treatment, compounds 9, 10, 11 and 12
showed a maximum effect after 48 h according to the MIC
values, suggesting that these molecules should be bio-
transformed to more active species. After 24 h of treat-
ment, nitro derivative 6 had a trichomonicidal activity
(MIC) comparable to that given by metronidazole. Like-
wise, nitro-ester 3 and the nitro-cyano derivatives 9 and 10
Acknowledgments Financial support through grant No. 49937,
´
Consejo Nacional de Ciencia y Tecnologıa (CONACYT) Mexico is
gratefully acknowledgedSpecial thanks to Dr. Marco Meraz,
´
´
123

420
Med Chem Res (2012) 21:415–420
´
Departamento Biologıa Molecular, CINVESTAV Mexico for kindly
providing the strains of Trichomonas vaginalis
´
antiparasitic activity of glutathione derivatives. Biomed Live Sci
387:303–305
Hansch C, Steward AR, Anderson SM, Bentley D (1968) The
parabolic dependence of drug action upon lipophilic character as
revealed by a study of hypnotics. J Med Chem 11(1):1–11
Leo A, Hansch C, Elkins D (1971) Partition coefficients and their
uses. Chem Rev 71:525–554
Lombardino JG (1965) Preparation and reactions of imidazo[1, 2-
a]pyridines. J Org Chem 30:2403–2407
Moridani MY, Siraki A, Chevaldina T, Scobie H, O’Brien PJ (2004)
Quantitative structure toxicity relationships for catechols in
isolated rat hepatocytes. Chem Biol Interact 147:297–307
References
ˇ
´
´
Acanski MM, Ðakovic-Sekulic TLj (2004) Correlation between
retention constants obtained in reversed-phase liquid chroma-
tography and partition coefficients of some benzimidazole
derivatives. APTEFF 35:1–280, BIBLID 1450–7188(35):165–
177
Arias L, Salgado-Zamora H, Campos E, Reyes A, Cervantes H,
Taylor EC (2006) Some nucleophilic substitutions in 2-cyano-3-
nitroimidazo[1, 2-a]pyridine. J Heterocycle Chem 43:565–569
Butler K, Howes HL, Lynch JE, Pirie DK (1967) Nitroimidazole
derivatives. Relationship between structure and antitrichomonal
activity. J Med Chem 10:891–897
Diamond LS, Graham CC, Cunnick CC (1995) YI-S, a casein-free
medium for anexic cultivation of Entamoeba hystolitica, related
YI-S, a casein-free medium for anexic cultivation of Entamoeba
hystolitica, related Entamoeba, Giardia intestinalis, Tricho-
monas vaginalis. J Eur Microbiol 42:277–278
´
Rizo B, Salgado H, Campos E, Jimenez R, Reyes A (2004) A
convenient synthesis of novel pyrido(1, 2:1, 2)imidazo[5,4-d]-
1,2,3-triazinones from imidazo[1, 2-a] pyridines. J Heterocyclic
41:91–94
Sangster J (1997) Octanol-water partition coefficients: fundamentals
and physical chemistry. Wiley, New York, pp 79–112
Teulade JC, Grassy G, Girard JP, Chapat JP, De Buochberg MS
´
(1978) Antibacteriens derives de la nitro-3-imidazo[1, 2-a]pyr-
idine. Synthese et relation structure-activite ‘‘in Vitro’’. Eur J
´
´
Med Chem 13:271–276
Winkelmann E, Raether W, Hartung H, Wagner WH (1977)
Chemotherapeutisch wirksame nitroverbindungen. 3. Mitteilung:
Nitropyridine, nitroimidazopyridine und vezwandte verbindu-
gen. Arzneim Forsch 27:82–89
Yajko MD, Madej JJ, Lancaster VM, Sanders AC, Hadley WK (1995)
Colorimetric method for determining MICs of antimicrobial
agents for Mycobacterium tuberculosis. J Clin Microbiol 33:
2324–2327
´
´
´
´
¨
Eros D, Kovesdi I, Orfi L, Takacs-Novak K, Acsady G, Keri G (2002)
Reliability of log P predictions based on calculated molecular
descriptors: a critical review. Curr Med Chem 9:1819–1829
Gensmantel NP (1994) Physicochemical properties and drug design.
In: King FD (ed) Medicinal chemistry principles and practice.
The Royal Society of Chemistry, Great Britain, pp 98–110
Golovanov IB, Zhenodarova SM, Ivanitskii GR, Khabarova MI
(2002) Structure-property correlation and estimation of the
123