• Nu S-Au Găsit Rezultate

(1)QSAR STUDY OF DISUBSTITUTED N6-CYCLOPENTYLADENINE ANALOGUES AS A ADENOSINE A1 RECEPTOR ANTAGONIST Abhishek K

N/A
N/A
Protected

Academic year: 2022

Share "(1)QSAR STUDY OF DISUBSTITUTED N6-CYCLOPENTYLADENINE ANALOGUES AS A ADENOSINE A1 RECEPTOR ANTAGONIST Abhishek K"

Copied!
11
0
0

Text complet

(1)

QSAR STUDY OF DISUBSTITUTED N6-CYCLOPENTYLADENINE ANALOGUES AS A ADENOSINE A1 RECEPTOR ANTAGONIST

Abhishek K. Jaina, V. Ravichandrana, Rajesh Singha, Simant Sharmaa, V. K. Mouryab, R. K. Agrawala*

aPharmaceutical Chemistry Research Laboratory, Department of Pharmaceutical Sciences, Dr. H. S. Gour Vishwavidyalaya, Sagar (M.P) – 470 003, India

bGovernment College of Pharmacy, Aurangabad (MH), India

In pursuit of better adenosine A1 receptor antagonist agents, QSAR studies were performed on a series of disubstituted N6-cyclopentyladenine analogues. Stepwise multiple linear regression analysis was performed to derive QSAR models which were further evaluated for statistical significance and predictive power by internal and external validation. The best QSAR model was selected, having correlation coefficient (r) = 0.879, standard error of estimation (SEE) = 0.368 and cross validated squared correlation coefficient (q2) = 0.664. The predictive ability of the selected model was also confirmed by leave one out cross validation. The QSAR model indicates that the dielectric energy, connectivity index 1, dipole vector Y, dipole vector Z, and HOMO energy play an important role for the A1 receptor antagonist activities. The results of the present study may be useful on the designing of more potent disubstituted N6-cyclopentyladenine analogues as adenosine A1 receptor antagonist agents.

(Received April 2, 2008; accepted April 9, 2008)

Keywords: QSAR; disubstituted N6-cyclopentyladenine analogues, Adenosine A1 receptor antagonist

1. Introduction

Denosine is a neuromotor which produces many important biological functions by activation of G protein coupled receptors that are classified in to A1 , A2B, and A3 subtypes.

Adenosine receptors from different species shows 87-93% amino acid sequence homology ,the only exception being the A3 subtypes which exhibit 74% primary sequence homology between rat and human [1-3] adenosine receptor are involved in many peripheral and central regulatory mechanism including vasodilation [4], vasoconstriction in the kidney [5], inhibition of lypolysis and insulin release [6] and moderation of cerebral ischemi [7].

The first A1 receptor antagonists were xantine derivatives , such as theophylline ,since then a variety of different classes of heterocyclic compounds has described to possess antagonist activity at adenosine receptor, xantine, adenines, 7-deazaadenine,7-deaza-8-ajapurine, pyrazolo (3- 4-c)quinolines, pyrazolo-(1-5-α) pyridine and 1-8-naphthyridine. E.W Van Tilburg synthesized a series of 4-methyl-(2-phenyl-carboxamido-)-1,3-thiazole derivatives as potential antagonist for the adenosine A1 receptors [8-14].

* Corresponding author: [email protected]

(2)

Rianne et al. [15] expressed N9 and C-8 position for increase adenosine A1 receptor affinity, small substituents at the 2-position of adenines or adenosines only have limited effects on adenosine A1 receptor affinity.

Linden and co-workers [16] investigated C-8 position of adenines to some extent. They synthesized 8-substituted N6-norbornyl-9-methyladenines and found that N-containig group at this position enhances A1 receptor affinity while introduction of alkyl chain on the C-8 position of adenines led to selective adenosine A3 receptor antagonist [17].

Computational chemistry has developed into an important contributor to rational drug design. Quantitative structure activity relationship (QSAR) modeling results in a quantitative correlation between chemical structure and biological activity. Senior author of the article Dr. R.

K. Agrawal and his research team has developed a few quantitative structure-activity relationship models to predict biological activity of different group of compounds [18-26].

2. Results and discussion

A data set of 37 compounds of reported series15 for adenosine A1 receptor antagonist activity was used for the present QSAR study (Table 1). The QSAR studies of the N6- cyclopentyladenine analogues series resulted in several QSAR equations. The two best equations are:

pKi = 10.653 (± 1.520) DE – 0.670 (± 0.256) CI1 +0.00311 (± 0.033) MR – 0.279 (± 0.106) DVZ + 0.450 (± 0.102) DVY + 2.782 (± 0.863) HE + 31.380 (± 8.420)……. (1)

n = 33, r = 0.883, r2 = 0.780, r2adj = 0.729, q2 = 0.638, F = 15.35, SEE = 0.3695, SPRESS = 0.430, P <

0.001.

pKi = 11.076 (± 1.451) DE – 0.434 (± 0.066) CI1 – 0.275 (± 0.106) DVZ + 0.508 (± 0.081) DVY + 3.224 (± 0.727) HE + 35.705 (± 7.080)……….. (2)

n = 33, r = 0.879, r2 = 0.772, r2adj = 0.730, q2 = 0.644, F = 18.30, SEE = 0.3689, SPRESS = 0.430, P <

0.001.

In the above equations n is the number of compounds used to derive the model and values in parentheses are the 95% confidence limit of respective coefficient. We extended our study for five-parametric correlations as they are permitted for a data set of 33 compounds in accordance with the lower limit of rule of thumb. Correlation matrix of the parameters in best model is given in table 3.

The calculated and predicted (LOO) activities of the compounds by the above models are shown in table 4. Model-1 shows good correlation coefficient (r) of 0.883 between descriptors (DE, CI1, MR, DVY, DVZ, and HE) and A1 receptor antagonist activity. Squared correlation coefficient (r2) of 0.780 explains 78.0% variance in biological activity.

This model also indicates statistical significance > 99.9% with F values F = 15.35. Cross validated squared correlation coefficient of this model was 0.638, which shows the good internal prediction power of this model. Model-2 shows good correlation coefficient (r) of 0.879 between descriptors (DE, CI1, HE, DVY, and DVZ) and A1 receptor antagonist activity. Squared correlation coefficient (r2) of 0.772 explains 77.2% variance in biological activity. This model also indicates statistical significance > 99.9% with F values F = 18.30. Cross validated squared correlation coefficient of this model was 0.644, which shows the good internal prediction power of this model.

Consequently equation-2 can be considered as most suitable model with both high statistical significant and excellent predictive ability.

The predictive ability of model-2 was also confirmed by external r2CVext. The robustness of the selected model was checked by Y – randomization test. The low r2 and q2 values indicate (data not shown) that the good results in our original model are not due to a chance correlation or structural dependency of the training set. The predictive ability of this model was also confirmed by external cross validation (equation 3). Consequently equation-2 can be considered as most suitable model with both high statistical significant and excellent predictive ability.

(3)

Table 1. Structures, biological activity of the N6-cyclopentyladenine analogues.

N

N N

N N H

R 8 9

N

N N

N N H

R 3

Compd. C8 N9-R N3-R Ki (nm),A1 receptor

1 2 3c

4 5 6 7 8 9 10 11 12 13 14 15

Br Br Br Br Br Br H Br Br Br OCH3

OCH3

OCH3

OC2H5

OCH(CH3)2

H Methyl Methyl Allyl Propyl Benzyl Benzyl

- - - Allyl Propyl Methyl Methyl Methyl

- - - - - - - Methyl

Propyl Benzyl

- - - - -

2646 43 467

35 33 1220 1810 2760 995 870 208 270 120 106 40

(4)

16 17 18 19 20

21

22

23

24

25

26

27

28

29

OC3H7

SC2H5

=O NHCH3

N

N

N

N

NHC2H5

NH

NH

N

N

N

Methyl Methyl Methyl

- -

- -

-

-

-

-

-

-

-

- - - - -

- -

-

-

-

-

-

-

-

235 224 1610

206 169

89 160

1011

344

2040

3560

7.7

5900

75

(5)

30

31

32

33

N

N

N

N O

-

-

-

-

-

-

-

-

706

28

68

2840

Ki = Displacement of {3H} DPCPX from CHO- A1 membrane

Table 2. Selected descriptors involved in developing QSAR models.

Comp. CI1 DVY DVZ DE HE MR

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

7.83 8.254 5.698 9.292 9.292 11.31 10.899

8.237 9.275 11.293

9.83 9.83 8.792 9.292 9.648 9.792 9.292

7.831 8.254 5.698 9.292 9.292 11.31 10.899 8.237 9.275 11.293 9.83 9.83 8.792 9.292 9.648 9.792 9.292

-0.529 -0.406 0.127 -0.795 -0.7 0.021 -0.075 -0.08 -0.025 0.163 -0.358 -0.42 -0.399 -0.485 -1.177 -0.536 -0.439

-0.61 -0.501 -0.707 -0.489 -0.471 -0.492 -0.528 -0.914 -0.836

-0.833 -0.447 -0.443 -0.468 -0.463 -0.451 -0.457 -0.451

-8.749 -8.693 -8.585 -8.685 -8.685 -8.689 -8.614 -8.593 -8.559 -8.527 -8.499 -8.507 -8.517 -8.507 -8.487 -8.504 -8.47

66.162 71.059 49.153 80.221 80.331 95.671 88.046 76.053 85.326 100.666 78.533 78.642 69.37 74.118 78.536 78.642 80.562

(6)

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

8.254 8.792 9.165 9.703 10.203 10.703 9.292 10.81 11.31 10.075 11.737 10.241 11.241 10.326 10.826 10.826

8.254 8.792 9.165 9.703 10.203 10.703 9.292 10.81 11.31 10.075 11.737 10.241 11.241 10.326 10.826 10.826

0.502 -1.451 -0.914 0.616 0.582 0.615 0.263 -1.5 -0.208 -1.977 -0.251 -1.276 -0.625 -1.265 -1.699 -0.688

-0.574 -0.557 -0.482 -0.442 -0.466 -0.455 -0.545 -0.528 -0.569 -0.454 -0.457 -0.437 -0.51 -0.466 -0.465 -0.543

-8.975 -8.398 -8.563 -8.534 -8.559 -8.482 -8.376 -8.394 -8.57 -8.534 -8.497 -8.429 -8.464 -8.494 -8.449 -8.513

64.748 73.101 77.336 82.084 86.608 91.209 77.849 89.513 91.393 86.502 96.29 86.832 95.957 84.876 89.477 86.410

CI1= Connectivity index order 1, DVY= Dipole vector Y, DVZ= Dipole vector Z, DE= Dielectric energy, HE = HOMO energy, MR= Molar refractivity.

Table 3. Correlation matrix between descriptors which are present in model.

BA CI1 DVY DVZ DE HE MR BA 1

CI1 -0.211 1

DVY -0.285 -0.143 1

DVZ -0.415 -0.102 0.300 1

DE 0.456 0.294 -0.888 -0.277 1

HE 0.134 0.342 -0.371 -0.344 0.187 1

MR -0.191 0.320 0.061 -0.063 0.108 0.339 1

(7)

Table 4. Observed, calculated and predicted (LOO) activity of derivatives.

Model-1 Model-2 Compd. No. Obs. Act.a

(p Ki) Cal. Act. Pred. Act. Cal. Act. Pred. Act 1

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

-3.423 -1.633 -2.669 -1.544 -1.518 -3.086 -3.258 -3.440 -2.998 -2.939 -2.318 -2.431 -2.079 -2.025 -1.602 -2.371 -2.350 -3.207 -2.314 -2.228 -1.950 -2.204 -3.005 -2.537 -3.310 -3.551 -0.886 -3.770 -1.875 -2.849 -2.605 -1.832 -3.453

-2.924 -1.682 -2.794 -1.733 -1.633 -2.916 -3.330 -3.337 -2.695 -3.317 -2.163 -2.232 -2.129 -2.287 -2.022 -2.457 -1.869 -3.173 -2.271 -2.113 -2.244 -2.597 -2.627 -2.845 -2.855 -4.218 -1.442 -2.973 -1.628 -3.133 -2.444 -2.045 -3.134

-2.735 -1.693 -2.902 -1.773 -1.662 -2.869 -3.340 -3.280 -2.544 -3.628 -2.133 -2.205 -2.140 -2.322 -2.062 -2.464 -1.785 -3.140 -2.261 -2.103 -2.321 -2.691 -2.517 -2.925 -2.749 -4.599 -1.623 -2.857 -1.592 -3.230 -2.427 -2.086 -3.068

-2.995 -1.722 -2.809 -1.792 -1.696 -2.985 -3.280 -3.356 -2.692 -3.277 -2.053 -2.142 -2.009 -2.200 -1.980 -2.39 -1.963 -3.101 -2.249 -2.158 -2.267 -2.648 -2.691 -2.843 -2.864 -4.235 -1.497 -2.959 -1.648 -3.250 -2.446 -2.015 -3.032

-2.865 -1.739 -2.929 -1.835 -1.732 -2.963 -3.282 -3.3101 -2.539 -3.541 -2.034 -2.125 -2.003 -2.211 -2.009 -2.399 -1.931 -3.015 -2.234 -2.154 -2.356 -2.740 -2.616 -2.924 -2.760 -4.623 -1.670 -2.843 -1.616 -3.318 -2.429 -2.049 -2.992

a All data represent mean values for at least two separate experiments. Obs. Act. - Observed activity, Cal. Act. – Calculated activity, Pred. Act. – Predicted activity by leave one out cross validation.

The predictive ability of model-2 was also confirmed by external r2CVext. The robustness of the selected model was checked by Y – randomization test. The low r2 and q2 values indicate (data not shown) that the good results in our original model are not due to a chance correlation or structural dependency of the training set.The predictive ability of this model was also confirmed by external cross validation (equation 3). The selected model was externally validated by randomly making training set of 27 compounds and test set of 6 compounds (28, 29, 30, 31, 32 and 33) (Table 5). QSAR was performed for training set and a model 3 was developed. This model was used to predict the biological activities of test set of compound.

pKi = 11.289 (± 1.609) DE – 0.391 (± 0.075) CI1 – 0.267 (± 0.116) DVZ + 0.500 (± 0.088) DVY + 3.166 (± 0.793) HE + 34.972 (± 7.850)………. (3)

n = 27, r = 0.884, r2 = 0.781, r2adj = 0.729, F = 14.96, SEE = 0.3632, P < 0.001.

(8)

Table 5. Predicted activity of test compounds.

Compd. Observed activity Predicted activity

28 29 30 31 32 33

-3.771 -1.875 -2.849 -2.605 -1.833 -3.453

-2.898 -1.563 -3.359 -2.531 -1.884 -2.939

The variables used in the selected model have no mutual correlation. This model showed good correlation coefficient (r) of 0.884 between descriptors Dielectric energy, Connectivity index 1, Diploe vector Y, Dipole vector Z and HOMO Energy and A1 receptor antagonist activity. Squared correlation coefficient (r2) of 0.781 explains 78.1% variance in biological activity.

The positive contribution of dielectric energy, dipole vector Y and HOMO energy on the biological activity showed that the increase in the values of these parameters lead to better A1- receptor antagonistic properties. The negative coefficient of connectivity index 1 indicated that the increase of CI1 is detrimental to biological activity and the negative coefficient of dipole vector Z is conducive to activity. Based on the developed QSAR model, new A1-receptor antagonist derivatives can be designed with caution.

The predicted activities of newly designed series (table 6) of compounds show that they all have predicted activities ranging from Ki (nm) = 0.57µM to 6.7 µM whereas the reported series has most active compound with Ki (nm) = 7.7 µM.

3. Experimental 3.1 General Procedure:

Win CAChe 6.1 (molecular modeling software, a product of Fujitsu private limited, Japan), Molecular modeling pro 6.1.0 (trial version, Cambridge software Corp.), STATISTICA version 6 (StatSoft, Inc., Tulsa, USA).

Table 6. The new designed series of compounds based on model 3

Comp.No. Compounds Structure Ki (nm),A1 receptor

1 N

N N

N N H

C H3 N H

6.7

2

N

N N

N N H

F

CH2-CH2-CH=CH2

4.78

(9)

3

N

N N

N N H

CF3 C H3

1.33

4

N

N N

N N H

CF3

CH2-CH2-CH=CH2

0.57

5

N

N N

N N H

B r

C H2- C H2- C H3

4.6

6

N

N N

N N H

CH2-CH3 NH

0.63

7

N

N N

N N H

CH3 O C H3

4.8

A data set of 33 compounds for A1-receptor antagonist activity was used for the present QSAR study. The molar concentrations of the compounds required to produce binding at receptor site (in nm) converted to free energy related negative logarithmic values for undertaking the QSAR study.

All 33 compounds’ structure were built on workspace of Win CAChe 6.1 (molecular modeling software, a product of Fujitsu private limited, Japan) and energy minimization of the molecules was done using Allinger’s MM2 force field followed by semi empirical PM3 method available in MOPAC module until the root mean square gradient value becomes smaller than 0.001 kcal/mol Å. Most stable structure for each compound was generated and used for calculating various physico-chemical descriptors like thermodynamic, steric and electronic values of descriptors.

3.2 Descriptors calculation, QSAR models development and validation

In present study the calculated descriptors were conformational minimum energies (CME), Zero-order connectivity index (CI0), First-order connectivity index (CI1), Second-order connectivity index (CI2), dipole moment (DM), total energy at its current geometry after optimization of structure (TE), heat of formation at its current geometry after optimization of structure (HF), highest occupied molecular orbital energies(HOMO), lowest unoccupied molecular orbital energies(LUMO), octanol-water partition coefficient(LOGP), molar refractivity(MR), shape index order 1 (SI1), shape index order 2 (SI2), shape index order 3 (SI3), Zero-order valance

(10)

connectivity index (VCI0), First-order valance connectivity index (VCI1), Second-order valance connectivity index (VCI2). Some of important descriptor which is present in model is shown in Table 2

All the calculated descriptors (50 descriptors calculated by Win CAChe 6.1 and Molecular modeling pro 6.1.0, the complete descriptors data set of all compounds will be provided on request) were considered as independent variable and biological activity as dependent variable.

STATISTICA version 6 (StatSoft, Inc., Tulsa, USA) software was used to generate QSAR models by stepwise multiple linear regression analysis. Statistical measures used were n-number of compounds in regression, r-correlation coefficient, r2-squared correlation coefficient, F- test (Fischer’s value) for statistical significance, SEE- standard error of estimation, q2- cross validated correlation coefficient and correlation matrix to show correlation among the parameters.

The squared correlation coefficient (or coefficient of multiple determination) r2 is a relative measure of fit by the regression equation. Correspondingly, it represents the part of the variation in the observed data that is explained by the regression. The correlation coefficient values closer to 1.0 represent the better fit of the regression. The F-test reflects the ratio of the variance explained by the model and the variance due to the error in the regression. High values of the F- test indicate that the model is statistically significant. Standard deviation is measured by the error mean square, which expresses the variation of the residuals or the variation about the regression line. Thus standard deviation is an absolute measure of quality of fit and should have a low value for the regression to be significant.

The predictive ability of the generated correlations was evaluated by cross validation method employing a ‘leave-one-out’ scheme. Validation parameters considered were cross validated r2 or q2, standard deviation based on predicted residual sum of squares (SPRESS) and standard error of prediction (SDEP). The predictive ability of the selected model was also confirmed by external r2CVext.

r

2

CVext = 1 -

test

i=1 test

i=1

(y

exp

- y

pred

)

2

(y

exp

- y

tr

)

2

The robustness of a QSAR model was checked by Y – randomization test. In this technique, new QSAR models were developed by shuffling the dependent variable vector randomly and keeping the original independent variable as such. The new QSAR models are expected to have low r2 and q2 values. If the opposite happens then an acceptable QSAR model can not be obtained for the specific modeling method and data.

Acknowledgement

One of the authors Mr. Abhishek Jain is grateful to U.G.C for providing fellowship for this work.

References

[1] C. A Salvatore, M. A Jacobson, H. E Taylor, J. Linden, R. Johnson, Proc. Natl. Acad. Sci.

U.S.A. 90, 10365 (1993).

[2] J. Linden, Trends Pharmacol. Sci. 15, 298 (1994).

[3] J. P. Hannon, H. J. Pfannkuche, J. R. Fozard, Br. J. Pharmacol 115, 945 (1995).

[4] R. A. Olsson, J. D. Pearson, Pharmacol. Rev. 3, 761 (1990).

[5] N. F. Rossi, P. C. Churchill, K. A. Jacoson, A. E. Leahy, Pharmacol. Exper. Ther. 240, 911 (1987).

[6] C. Londson, D. M. Cooper, Woff, J.Prac.Natl.Acad.sci. USA 77, 2251 (1980).

[7] B. D. Hillaire, G. Bertrand, R. Gross, Eur.J.Pharmacol. 136, 109 (1987).

(11)

[8] L. Betti, G. Biagi, G. Giannaccini, I. Giorgi, O. Livi, A. Lucacchini, C. Manera, V. Scartoni, J.

Med. Chem. 41, 668 (1998).

[9] E. Camaioni, S. Costanzi, S. Vittori, R. Volpini, K. N. Klotz, G. Cristalli, Bioorg. Med. Chem.

6, 523 (1998).

[10] V. Collota, D. Catarzi, F. Varano, L. Cecchi, G. Filacchioni, C. Martini, L. Trincavelli, A.

Lucacchini, J. Med. Chem. 43, 3118 (2000).

[11] P. L. Ferrarini, C. Mori, C. Manera, A. Martinelli, F. Mori, G. Saccomanni, P. L. Barili, L.

Betti, G. Giannaccini, L. Trincavelli, A. Lucacchini, J. Med. Chem. 43, 2814 (2000).

[12] T. Katsushima, L. Nieves, N. Wells, J. Med. Chem. 33, 1906 (1990).

[13] S. Kuroda, A. Akahane, H. Itani, S. Nishimura, K. Durkin, T. Kinoshita, Y. Tenda, K.

Sakane, Bioorg. Med. Chem. Lett. 9, 1979 (1999).

[14] C. E. Muller, Expert Opin. Ther. Pat. 5, 419 (1997).

[15] A. F. Rianne, D. Ligt, A. M. Pieter, V. D. Klein, K. Jacobien, F. Drabbe Kunzel, A. Lorenzen, F. A. Maate, S. Fujikawa, R. V. Westhoven, T. V. D. Hoven, J. Brussee, P. Ijzerman, Bioorg.

Med. Chem.12, 139 (2004).

[16] P. L. Martin, R. J. Wysocki, R. J. Barrett, J. M. May, J. Linden, J. Pharmacol. Exp. Ther. 276, 490 (1996).

[17] R. Volpini, S. Costanzi, C. Lambertucci, S. Vittori, A. Lorenzen, K. N. Klotz, G. Cristalli, Bioorg. Med. Chem. Lett. 11, 1931 (2001).

[18] V. Ravichandran, R. K. Agrawal, Bioorg. Med. Chem. Lett. 17, 2197 (2007).

[19] V. Ravichandran, V. K. Mourya, R. K. Agrawal, Arkivoc XIV, 204 (2007).

[20] K. K. Sahu, V. Ravichandran, V. K. Mourya, R. K. Agrawal, Med. Chem. Res. 15, 418 (2007).

[21] K. K. Sahu, V. Ravichandran, P. K. Jain, S. Sharma, V. K. Mourya, R. K. Agrawal, Acta Chim. Slov. (Inpress) (2007).

[22] S. Sharma, V. Ravichandran, P. K. Jain, V. K. Mourya, R. K. Agrawal, J. Enzym. Inhib. Med.

Chem. (Inpress) (2007).

[23] H. K. Jain, R. K. Agrawal, Inter. Elect. J. Mol. Design 5, 224 (2006).

[24] H. K. Jain, R. K. Agrawal, Inter. Elect. J. Mol. Design (Inpress) (2007).

[25] H. K. Jain, V. K. Mourya, R. K. Agrawal, Bioorg. Med. Chem. Lett. 16, 5280 (2006).

[26] S. J. Dighade, H. K. Jain, R. K. Agrawal, Indian J. Pharm. Sci. 65, 586 (2003).

Referințe

DOCUMENTE SIMILARE

Our data also showed a moderate correlation be- tween 2D-SWE.GE and 2D-SWE.SSI (r=0.63, R 2 =0.4, p&lt;0.0001), results which could be accountable to the nar- row range

Identity is thus constructed in interaction, which means that out of a whole host of potential identity features, those features become salient which permit a differentiation of

We here by report the models, as generated by kNN-MFA in conjunction with simulated annealing (SA), genetic algorithm (GA), and stepwise (SW) forward variable selection methods

The evolution to globalization has been facilitated and amplified by a series of factors: capitals movements arising from the need of covering the external

The methodology of managerial training programs’ development at university level, and, particularly, at post-university level, goes on, with few exceptions, to be

The specific legislation regarding cosmetics (Table 2) settles the main problems concerning the manufacturing, commercialization, goods safety in usage and customer’s security..

In order to describe the economic implications of a prolonged military rivalry, we have constructed a nonlinear dynamical model that merges the classical Richardson arms race

De¸si ˆın ambele cazuri de mai sus (S ¸si S ′ ) algoritmul Perceptron g˘ ase¸ste un separator liniar pentru datele de intrare, acest fapt nu este garantat ˆın gazul general,