English

• 1. INTRODUCTION

  Alzheimer's disease (AD) is a typical representative of neurodegenerative diseases. It has become the most common cause of dementia in the elderly, and it is also a common cause of disability in individuals over 65 years of age^[1]. AD is a very complex neurological disease that is affected by many factors. Over the years, in order to understand the pathogenesis of AD, scientists have proposed a variety of hypotheses, including the cholinergic damage hypothesis^[2], Tau protein hypothesis^[3], metal ion homeostasis imbalance hypothesis^[4], β-amyloid protein cascade hypothesis^[5], neuro-inflammation hypothesis^[6], genetic inheritance hypothesis^[7], oxidative stress hypothesis^[8] and so on. The most typical of them is the central cholinergic damage hypo- thesis, which considers acetylcholine (ACh) as an important chemical substance in the nervous system. When AD disease occurs, the synthesis, storage and release of ACh are reduced by the reduction of cholinergic neurons in the brain. This leads to major clinical manifestations such as memory degradation and recognition of dysfunction. Therefore, maintaining and restoring Ach levels can alleviate the symptoms of AD disease^{[9, 10]}. Acetylcholine inhibitors (AChi) play a role in the treatment of AD by inhibiting the activity of AChE, delaying the rate of ACh hydrolysis, and increasing the level of ACh in the synaptic cleft^[11]. Thus, AChEIs are involved in a wide range of AD drug discovery studies^[12].

  Chromenone is a chemical compound which has a benzene ring and a six-membered heterocyclic compound with an oxygen hetero atom. Chromenones themselves have low value for treating AD, but some of their derivatives have been found to be key scaffolds for optimal AChEI activity in the interaction with CAS and PAS of AChE. In recent years, chromenon derivatives have been the most promising AChEi^[13]. Apart from chromenones, 1, 2, 3-trrizoles have been considered as versatile anti-AChE agents^[14]. In recent years, it has been reported that many scientists have focused on the combination of the two structures, hoping to obtain excellent AChEi.

  Therefore, we retrieved a series of new 1, 2, 3-triazole-chromenone derivatives according to the Arezoo Rastegari team and Mina Saeedi team^{[15, 16]}, and recorded the activity of them as AChEi. we attempted to illuminate the structure-activity relationships (SARs) to provide useful guidelines for the design of new 1, 2, 3-triazole-chromenone derivatives as potent AChEIs, which are based on the established 3D-QSAR comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) models for a series of disclosed triazole-chromenone analogues.

  2. MATERIALS AND METHODS

  2.1 Preparation of biologically active molecular data sets

  The datasets of 25 1, 2, 3-triazole-chromenone derivatives for 3D-QSAR analysis were derived from two papers^{[15, 16]} and the molecular activity assays were consistent. First, we convert the half maximal inhibitory concentration (IC₅₀) values to the pIC₅₀ (–log₁₀(IC₅₀)) values. The obtained pIC₅₀ values ranging between 3.989 and 5.745 were used as response variables for subsequent 3D-QSAR analyzes. The 25 molecules were randomly divided into training sets (20) to construct CoMFA and CoMSIA models and test sets (5) to test the credibility of the model. All compound structures and their IC₅₀ are listed in Table 1. The 3D structures of 1, 2, 3-triazole-chromenone derivatives were established using standard geometric parameters of molecular modeling software package SYBYL-X 2.0. At first, all compounds must be cleaned and optimized. We used the Powell conjugate gradient algorithm with a convergence criterion of 0.01 kcal/mol Å to perform energy minimization in the Tripos force field with a distance-dependent dielectric. Partial atomic charges were calculated by the Gasteiger-Hückel method^[17-20].

  Table 1

  

  Table 1.  Structures and Bioactivity Data of the Compounds in the Dataset

  DownLoad: CSV

  Sr.no	Compound numbers	X	Ar	Experimental pIC50	Predicted pIC50
  					CoMFA	Residue	CoMSIA	Residue
  1	1a	-	H	5.347	5.299	–0.048	5.339	–0.008
  2**	1b	-	3, 4-Me	5.745	5.732	–0.013	5.817	0.072
  3	1c	-	2-Me	5.432	5.442	0.010	5.404	–0.028
  4	1d	-	2-Br	5.483	5.521	0.038	5.549	0.066
  5	1e	-	2-Cl	5.481	5.497	0.016	5.393	–0.088
  6	1f	-	3, 4-Cl	5.690	5.695	0.005	5.657	–0.033
  7	1g	-	4-F	4.996	4.992	–0.004	5.013	0.017
  8	2a	-	4-Me	4.361	4.356	–0.005	4.447	0.086
  9	2b	-	2-Me	5.007	5.013	0.006	5.043	0.036
  10	2c	-	2-Br	5.678	5.605	–0.073	5.584	–0.094
  11	2d	-	2, 3-Cl	5.298	5.360	0.062	5.398	0.100
  12*	2e	-	3, 4-Cl	3.989	4.700	0.711	4.576	0.587
  13*	2f	-	4-Cl	4.532	4.676	0, 144	4.591	0.059
  14	2g	-	3-Cl	4.790	4.722	–0.068	4.666	–0.124
  15	2h	-	4-F	4.790	4.827	0.037	4.776	–0.014
  16	2i	-	3-F	4.648	4.684	0.036	4.662	0.014
  17*	3a	H	3-F	4.318	4.758	0.440	4.716	0.398
  18	3b	H	2-Cl	4.469	4.449	–0.020	4.459	–0.010
  19*	3c	H	2, 3-Cl	4.694	4.730	0.036	4.694	0.000
  20	3d	H	3, 4-Cl	4.777	4.806	0.029	4.754	–0.023
  21*	3e	H	2-Br	4.508	4.635	0.127	4.601	0.093
  22	3f	OMe	2-Me	4.618	4.625	0.007	4.688	0.070
  23	3g	OMe	2-Cl	4.812	4.805	–0.007	4.758	–0.054
  24	3h	OMe	2, 3-Cl	4.472	4.501	0.029	4.553	0.081
  25	3i	OMe	3, 4-Cl	4.593	4.556	–0.037	4.528	–0.065
  * Represents molecules in the test set; **Represents the most active inhibitor; CoMFA, comparative molecular field analysis; CoMSIA, comparative molecular similarity indices analysis

  2.2 Molecular modelling and alignment

  The 3D structure alignment is an important factor that significantly affects the results of 3D-QSAR studies. Various different alignment strategies have been well described in the literature^{[21, 22]}. The alignment rules determine the quality and predictability of the established model. Typically, one of the following is used as the template molecule: (a) the lead and/or commercial compound, (b) the most active compound, and (c) the compound having the most functional groups. In this study, we obtained the global minimum-energy conformation of the most active inhibitor 2 and used it as a template molecule, which was then aligned with the rigid body of ALIGN DATABASE to align the geometrically optimized molecules. Fig. 1 shows the structure of compound 2 with the largest common substructure. It can be observed that all compounds in this study have similar active conformations (Fig. 2).

  Figure 1

  

  Figure 1.  Public skeleton structure

  DownLoad: Full-Size Img PowerPoint

  Figure 2

  

  Figure 2.  Alignment of the molecules in the dataset

  DownLoad: Full-Size Img PowerPoint

  2.3 Partial least-squares analysis

  To calculate the size and distribution of the steric and electrostatic fields at each grid point around the overlapping molecules, we decided to use the Tripos force field. At the same time, we use sp³ carbon having a +1 charge as the probe atom, and the rest of the parameters were the system defaults^[23-25]. In partial least-squares (PLS) analysis, firstly, the leave-one-out (LOO) was used for cross-validation to obtain the optimal number of components (ONC). After determining the optimal number of components, non-cross-validation analysis was performed using column filtration. The cross-validation correlation coefficient Q², predicted residual sum of squares (PRESS), non-cross-validated correlation coefficient r_ncv² and standard standard error (SEE) value were calculated. The correlation coefficient of cross-validation was also calculated using the following equation^[26]:

  $ \mathrm{Q}^2=1-\left[\Sigma\left(\mathrm{y}_{\text {predicted }}-\mathrm{y}_{\text {observed }}\right)^2 / \Sigma\left(\mathrm{y}_{\text {observed }}-\mathrm{y}_{\text {mean }}\right)^2\right]$

  Where y_predicted, y_observed and y_mean are the predicted, observed, and mean activity values, respectively. After calculating Q², the calculation of r_pred² is also indispensable, because external verification using data outside the training set is considered to be the only method to guarantee the prediction ability of the QSAR model. The calculation formula of r_pred² is as follows^[27]:

  $\mathrm{r}_{\text {pred }}{ }^2=1-(\mathrm{PRESS} / \mathrm{SD}) $

  Where SD represents the sum of the squared deviations between the biological activity of the test group compound and the average activity of the training one, and PRESS is the sum of the squared deviations between the experimental activity and the predicted activity of the test group compound, respectively.

  3. RESULTS AND DISCUSSION

  3.1 Test the reliability of the model

  PLS analysis results of the CoMFA and CoMSIA are listed in Table 2. It is reported that the model is excellent and has good predictive power if Q² is greater than 0.5 and r_pred² is higher than 0.6^{[28, 29]}. In this study, the calculated Q² is 0.597 for CoMFA model and 0.721 for CoMSIA model, and the calculated r_pred² for the CoMFA and CoMSIA models is 0.728 and 0.805, respectively. These results indicate that both these two models have good predictive power. In order to further verify the predictability of these two models, we use the model trained by the training set to predict the test set. Fig. 3 shows that the linear correlation between the experimental and predicted activities of all molecules is very great. The residue between the predicted and experimental pIC₅₀ values is within the tolerable range, and the correlation between the two values is also great, which further verified that both CoMFA and CoMSIA have excellent predictive power.

  Table 2

  

  Table 2.  Summary of Partial Least-squares (PLS) Results

  DownLoad: CSV

  Parameters	PLS results
  	CoMFA	CoMSIA
  Q2 (cross-validated)	0.597	0.721
  rpred2 (test)	0.728	0.805
  PRESS	0.737	0.528
  SD	2.708	2.708
  Standard error of estimate	0.043	0.076
  Optimum number of components	6	5
  rncv2 (non-cross-validated)	0.994	0.979
  Test value	346.726	131.107
  Steric field contribution	0.609	0.148
  Electrostatic field contribution	0.391	0.261
  Hydrophobic field contribution	-	0.569
  Hydrogen bond donor field contribution	-	0.021
  Hydrogen bond acceptor field contribution	-	0

  Figure 3

  

  Figure 3.  Scatterplot of experimental versus predicted pIC₅₀ by comparative molecular field analysis (CoMFA) (a) and comparative molecular similarity indices analysis (CoMSIA) (b)

  DownLoad: Full-Size Img PowerPoint

  3.2 3D-QSAR contour maps

  The field distribution of the models could be seen from the 3D-QSAR contour maps. The steric and electrostatic contour maps around reference molecule 2 of the CoMFA and CoMSIA models are shown in Figs. 4, 5(a) and 5(b). CoMFA and CoMSIA steric contour maps are displayed in Fig. 4(a) and 5(a). The green color block shows that increasingly the steric hindrance of substituent in this region can enhance the molecular activity, and the yellow color block means that reducing the steric hindrance of the substituent in this region can enhance the molecular activity. For example, molecule 2 exhibits higher anti-AD activity than molecule 3, mainly because it exhibited a larger steric hindrance than 3 at the 3rd position of the benzene ring. This means that the introduction of bulky groups at this position will be beneficial for anti-AD efficacy. CoMFA and CoMSIA electrostatic contour maps are represented in Fig. 4(b) and 5(b). Blue and red represent favorable and unfavorable regions for electrostatic field. As shown in Fig. 4(b) and 5(b), the 2nd and 3rd positions on the benzene ring are surrounded by red color blocks, and the blue color blocks are wrapped around the 4th position on the benzene ring, which indicates that the addition of negatively charged groups in the 2nd and 3rd positions favors molecular activity, but the 4th position is exactly opposite. For example, under the same conditions of basic skeleton, the activity of molecule 11 is significantly greater than the activity of molecule 12. The reason for this phenomenon is because the "Cl" atoms in the 2nd and 3rd positions of compound 11 are negatively charged, while the 4th position of molecule 12 has a "Cl", which increases the negative charge of the 4th position on the benzene ring. Hydrogen bond donors and acceptors have the lowest contribution to molecular activity, so we omitted the analysis of their contour map.

  Figure 4

  

  Figure 4.  Contour maps of the CoMFA models

  DownLoad: Full-Size Img PowerPoint

  Figure 5

  

  Figure 5.  Contour maps of the CoMSIA models

  DownLoad: Full-Size Img PowerPoint

  4. DESIGN OF NEW DRUGS

  Fig. 6 is a predicted molecular template. According to Table 2, the contribution of the steric field is greater than that of the electrostatic field for both the CoMFA and CoMSIA models, which indicates that the activity of the drug molecule is mainly affected by the size and steric hindrance of the substituent, so we designed the N1 (CoMFA: 5.962, CoMSIA: 5.542) and N2 (CoMFA: 5.743, CoMSIA: 5.841) molecules. They replace the methyl group at the R3 position with a methoxy group and an ethyl group respectively, and their pharmacological activity was predicted based on the model. The predicted activity of the two molecules is significantly increased by comparison with the activity of the most active compound 2. Besides the steric field, the hydrophobicity also plays a large role in the CoMSIA model. According to Fig. 5(c), replacing a hydrophobic group at the R3 position can increase the molecular activity, and then based on this 3D-QSAR hydrophobic field we designed N3 (CoMFA: 5.801, CoMSIA: 5.890) and N4 (CoMFA: 5.770, CoMSIA: 6.064).

  Figure 6

  

  Figure 6.  Template for predicting the molecules

  DownLoad: Full-Size Img PowerPoint

  Table 3

  

  Table 3.  Predictive Activity of the Designed Small Molecules Based on CoMFA and CoMSIA Models

  DownLoad: CSV

  Compound_ ID	Compound_ structure	CoMFA	CoMSIA
  N1		5.962	5.542
  N2		5.743	5.841
  N3		5.801	5.890
  N4		5.760	6.064

  5. CONCLUSION

  In this study, 25 structurally related 1, 2, 3-triazole-chromenone derivatives from the literature were selected for 3D-QSAR analysis. Through the analysis of the 3D-QSAR contour maps generated by SYBYL-X 2.0 software, the main relevant structural conclusions of this study are as follows: (i) The large population of R3 in Fig. 6 is more favorable for anti-AD efficacy than the small population; (ii) the compound having a halogen group at the R3 position is more active than the derivative without it in Fig. 6. Then, the cross-validation was conducted to obtain high predictive and satisfactory CoMFA model (Q² = 0.597, R² = 0.994) and CoMSIA model (Q² = 0.721, R² = 0.979), the essential external verification (r_pred² = 0.728, 0.805 for CoMFA and CoMSIA) shows that the model has significant predictive effects.

  
  
• 1. [1]

     Bolos, M.; Perea, J. R.; Avila, J. Alzheimer's disease as an inflammatory disease. Biomol. Concepts 2017, 8, 37–43. doi: 10.1515/bmc-2016-0029

  2. [2]

     Bartus, R. T.; Dean, R. L.; Beer, B.; Lippa, A. S. The cholinergic hypothesis of geriatric memory dysfunction. Science (New York, N. Y. ) 1982, 217, 408–14. doi: 10.1126/science.7046051

  3. [3]

     Iqbal, K.; Alonso, A. D. C.; Chen, S.; Chohan, M. O.; El-Akkad, E.; Gong, C. X.; Khatoon, S.; Li, B.; Liu, F.; Rahman, A.; Tanimukai, H.; Grundke-Iqbal, I. Tau pathology in Alzheimer disease and other tauopathies. BBA-Molecular Basis of Disease 2005, 1739, 198–210. doi: 10.1016/j.bbadis.2004.09.008

  4. [4]

     Bush, A. I. Metals and neuroscience. Curr. Opin. Chem. Biol. 2000, 4, 184–191. doi: 10.1016/S1367-5931(99)00073-3

  5. [5]

     Hardy, J. A.; Higgins, G. A. Alzheimer's disease: the amyloid cascade hypothesis. Science (New York, N. Y. ) 1992, 256, 184–5. doi: 10.1126/science.1566067

  6. [6]

     Finch, C. E.; Morgan, T. E. Systemic inflammation, infection, apoE alleles, and Alzheimer disease: a position paper. Curr. Alzheimer Res. 2007, 4, 185–189. doi: 10.2174/156720507780362254

  7. [7]

     Bird, T. D. Genetic factors in Alzheimer's disease. N. Engl. J. Med. 2005, 352, 862–864. doi: 10.1056/NEJMp058027

  8. [8]

     Pratico, D. Oxidative stress hypothesis in Alzheimer's disease: a reappraisal. Trends Pharmacol. Sci. 2008, 29, 609–615. doi: 10.1016/j.tips.2008.09.001

  9. [9]

     Xie, Q.; Wang, H.; Xia, Z.; Lu, M.; Zhang, W.; Wang, X.; Fu, W.; Tang, Y.; Sheng, W.; Li, W.; Zhou, W.; Zhu, X.; Qiu, Z.; Chen, H. Bis-(-)-nor-meptazinols as novel nanomolar cholinesterase inhibitors with high inhibitory potency on amyloid-beta aggregation. J. Med. Chem. 2008, 51, 2027–2036. doi: 10.1021/jm070154q

  10. [10]

      Munoz-Torrero, D. Acetylcholinesterase inhibitors as disease-modifying therapies for alzheimer's disease. Curr. Med. Chem. 2008, 15, 2433–2455. doi: 10.2174/092986708785909067

  11. [11]

      Bartolucci, C.; Haller, L. A.; Jordis, U.; Fels, G.; Lamba, D. Probing torpedo californica acetylcholinesterase catalytic gorge with two novel bis-functional galanthamine derivatives. J. Med. Chem. 2010, 53, 745–751. doi: 10.1021/jm901296p

  12. [12]

      Pohanka, M. Acetylcholinesterase inhibitors: a patent review (2008-present). Expert Opin. Ther. Pat. 2012, 22, 871–886. doi: 10.1517/13543776.2012.701620

  13. [13]

      Anand, P.; Singh, B.; Singh, N. A review on coumarins as acetylcholinesterase inhibitors for Alzheimer's disease. Bioorg. Med. Chem. 2012, 20, 1175–1180. doi: 10.1016/j.bmc.2011.12.042

  14. [14]

      Lewis, W. G.; Green, L. G.; Grynszpan, F.; Radic, Z.; Carlier, P. R.; Taylor, P.; Finn, M. G.; Sharpless, K. B. Click chemistry in situ: acetylcholinesterase as a reaction vessel for the selective assembly of a femtomolar inhibitor from an array of building blocks. Angew. Chem. Int. Ed. 2002, 41, 1053. doi: 10.1002/1521-3773(20020315)41:6<1053::AID-ANIE1053>3.0.CO;2-4

  15. [15]

      Rastegari, A.; Nadri, H.; Mahdavi, M.; Moradi, A.; Mirfazli, S. S.; Edraki, N.; Moghadam, F. H.; Larijani, B.; Akbarzadeh, T.; Saeedi, M. Design, synthesis and anti-Alzheimer's activity of novel 1, 2, 3-triazole-chromenone carboxamide derivatives. Bioorg. Chem. 2019, 83, 391–401. doi: 10.1016/j.bioorg.2018.10.065

  16. [16]

      Saeedi, M.; Safavi, M.; Karimpour-Razkenari, E.; Mahdavi, M.; Edraki, N.; Moghadam, F. H.; Khanavi, M.; Akbarzadeh, T. Synthesis of novel chromenones linked to 1, 2, 3-triazole ring system: investigation of biological activities against Alzheimer's disease. Bioorg. Chem. 2017, 70, 86–93. doi: 10.1016/j.bioorg.2016.11.011

  17. [17]

      Dixon, S. L.; Smondyrev, A. M.; Knoll, E. H.; Rao, S. N.; Shaw, D. E.; Friesner, R. A. PHASE: a new engine for pharmacophore perception, 3D QSAR model development, and 3D database screening: 1. Methodology and preliminary results. J. Comput. Aided Mol. Des. 2006, 20, 647–671. doi: 10.1007/s10822-006-9087-6

  18. [18]

      Fang, Y.; Lu, Y.; Zang, X.; Wu, T.; Qi, X.; Pan, S.; Xu, X. 3D-QSAR and docking studies of flavonoids as potent Escherichia coli inhibitors. Sci. Rep. 2016, 6, 23634. doi: 10.1038/srep23634

  19. [19]

      Liu, X. H.; Xu, X. Y.; Tan, C. X.; Weng, J. Q.; Xin, J. H.; Chen, J. Synthesis, crystal structure, herbicidal activities and 3D-QSAR study of some novel 1, 2, 4-triazolo 4, 3-a pyridine derivatives. Pest Manage. Sci. 2015, 71, 292–301. doi: 10.1002/ps.3804

  20. [20]

      Yang, Q.; Zhang, S. P.; Zhao, S. P. 3D-QSAR studies on a series of indoleamide derivatives as antiplasmodial drugs. Chin. J. Struct. Chem. 2018, 37, 1015–1024.

  21. [21]

      Castilho, M. S.; Postigo, M. P.; de Paula, C. B. V.; Montanari, C. A.; Oliva, G.; Andricopulo, A. D. Two- and three-dimensional quantitative structure-activity relationships for a series of purine nucleoside phosphorylase inhibitors. Bioorg. Med. Chem. 2006, 14, 516–527. doi: 10.1016/j.bmc.2005.08.055

  22. [22]

      Salum, L. D. B.; Polikarpov, I.; Andricopulo, A. D. Structural and chemical basis for enhanced affinity and potency for a large series of estrogen receptor ligands: 2D and 3D QSAR studies. J. Mol. Graphics Model. 2007, 26, 434–442. doi: 10.1016/j.jmgm.2007.02.001

  23. [23]

      Buolamwini, J. K.; Assefa, H. CoMFA and CoMSIA 3D QSAR and docking studies on conformationally-restrained cinnamoyl HIV-1 integrase inhibitors: exploration of a binding mode at the active site. J. Med. Chem. 2002, 45, 841–852. doi: 10.1021/jm010399h

  24. [24]

      Doytchinova, I. A.; Flower, D. R. Toward the quantitative prediction of T-cell epitopes: CoMFA and CoMSIA studies of peptides with affinity for the class I MHC molecule HLA-A*0201. J. Med. Chem. 2001, 44, 3572–3581. doi: 10.1021/jm010021j

  25. [25]

      Murthy, V. S.; Kulkarni, V. M. 3D-QSAR CoMFA and CoMSIA on protein tyrosine phosphatase 1B inhibitors. Bioorg. Med. Chem. 2002, 10, 2267–2282. doi: 10.1016/S0968-0896(02)00056-1

  26. [26]

      Awasthi, M.; Singh, S.; Pandey, V. P.; Dwivedi, U. N. CoMFA and CoMSIA-based designing of resveratrol derivatives as amyloid-beta aggregation inhibitors against Alzheimer's disease. Med. Chem. Res. 2018, 27, 1167–1185. doi: 10.1007/s00044-018-2138-4

  27. [27]

      Benigni, R.; Bossa, C. Predictivity of QSAR. J. Chem. Inf. Model. 2008, 48, 971–980. doi: 10.1021/ci8000088

  28. [28]

      Golbraikh, A.; Tropsha, A. Beware of q(2)! J. Mol. Graphics Modell. 2002, 20, 269–276. doi: 10.1016/S1093-3263(01)00123-1

  29. [29]

      Eriksson, L.; Jaworska, J.; Worth, A. P.; Cronin, M. T. D.; McDowell, R. M.; Gramatica, P. Methods for reliability and uncertainty assessment and for applicability evaluations of classification- and regression-based QSARs. Environ. Health Perspect. 2003, 111, 1361–1375. doi: 10.1289/ehp.5758

• 

  Figure 1  Public skeleton structure

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Alignment of the molecules in the dataset

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Scatterplot of experimental versus predicted pIC₅₀ by comparative molecular field analysis (CoMFA) (a) and comparative molecular similarity indices analysis (CoMSIA) (b)

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Contour maps of the CoMFA models

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Contour maps of the CoMSIA models

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Template for predicting the molecules

  下载: 全尺寸图片 幻灯片

  Table 1.  Structures and Bioactivity Data of the Compounds in the Dataset

  Sr.no	Compound numbers	X	Ar	Experimental pIC50	Predicted pIC50
  					CoMFA	Residue	CoMSIA	Residue
  1	1a	-	H	5.347	5.299	–0.048	5.339	–0.008
  2**	1b	-	3, 4-Me	5.745	5.732	–0.013	5.817	0.072
  3	1c	-	2-Me	5.432	5.442	0.010	5.404	–0.028
  4	1d	-	2-Br	5.483	5.521	0.038	5.549	0.066
  5	1e	-	2-Cl	5.481	5.497	0.016	5.393	–0.088
  6	1f	-	3, 4-Cl	5.690	5.695	0.005	5.657	–0.033
  7	1g	-	4-F	4.996	4.992	–0.004	5.013	0.017
  8	2a	-	4-Me	4.361	4.356	–0.005	4.447	0.086
  9	2b	-	2-Me	5.007	5.013	0.006	5.043	0.036
  10	2c	-	2-Br	5.678	5.605	–0.073	5.584	–0.094
  11	2d	-	2, 3-Cl	5.298	5.360	0.062	5.398	0.100
  12*	2e	-	3, 4-Cl	3.989	4.700	0.711	4.576	0.587
  13*	2f	-	4-Cl	4.532	4.676	0, 144	4.591	0.059
  14	2g	-	3-Cl	4.790	4.722	–0.068	4.666	–0.124
  15	2h	-	4-F	4.790	4.827	0.037	4.776	–0.014
  16	2i	-	3-F	4.648	4.684	0.036	4.662	0.014
  17*	3a	H	3-F	4.318	4.758	0.440	4.716	0.398
  18	3b	H	2-Cl	4.469	4.449	–0.020	4.459	–0.010
  19*	3c	H	2, 3-Cl	4.694	4.730	0.036	4.694	0.000
  20	3d	H	3, 4-Cl	4.777	4.806	0.029	4.754	–0.023
  21*	3e	H	2-Br	4.508	4.635	0.127	4.601	0.093
  22	3f	OMe	2-Me	4.618	4.625	0.007	4.688	0.070
  23	3g	OMe	2-Cl	4.812	4.805	–0.007	4.758	–0.054
  24	3h	OMe	2, 3-Cl	4.472	4.501	0.029	4.553	0.081
  25	3i	OMe	3, 4-Cl	4.593	4.556	–0.037	4.528	–0.065
  * Represents molecules in the test set; **Represents the most active inhibitor; CoMFA, comparative molecular field analysis; CoMSIA, comparative molecular similarity indices analysis

  下载: 导出CSV

  Table 2.  Summary of Partial Least-squares (PLS) Results

  Parameters	PLS results
  	CoMFA	CoMSIA
  Q2 (cross-validated)	0.597	0.721
  rpred2 (test)	0.728	0.805
  PRESS	0.737	0.528
  SD	2.708	2.708
  Standard error of estimate	0.043	0.076
  Optimum number of components	6	5
  rncv2 (non-cross-validated)	0.994	0.979
  Test value	346.726	131.107
  Steric field contribution	0.609	0.148
  Electrostatic field contribution	0.391	0.261
  Hydrophobic field contribution	-	0.569
  Hydrogen bond donor field contribution	-	0.021
  Hydrogen bond acceptor field contribution	-	0

  下载: 导出CSV

  Table 3.  Predictive Activity of the Designed Small Molecules Based on CoMFA and CoMSIA Models

  Compound_ ID	Compound_ structure	CoMFA	CoMSIA
  N1		5.962	5.542
  N2		5.743	5.841
  N3		5.801	5.890
  N4		5.760	6.064

  下载: 导出CSV
•