Available online on 15.03.2025 at http://jddtonline.info

Journal of Drug Delivery and Therapeutics

Open Access to Pharmaceutical and Medical Research

Copyright  © 2025 The  Author(s): This is an open-access article distributed under the terms of the CC BY-NC 4.0 which permits unrestricted use, distribution, and reproduction in any medium for non-commercial use provided the original author and source are credited

Open Access Full Text Article  Research Article

Identification of Some α-Glucosidase Inhibitors Using QSAR Modeling Based Drug Repurposing Approach 

Sonu ¹, Mohan Lal Kori ²*

¹ Ph. D. Research scholar RKDF University Bhopal (M.P.), India

² Vice Chancellor, Tantya Bhil University Khargone (M.P.), India

 

 

1. INDRODUCTION

Diabetes is a group of metabolic diseases characterized by hyperglycemia caused by inadequate insulin secretion with or without a simultaneous decrease in hormone action at its receptor ¹. 

Currently, diabetes is the fifth deadliest disease. As per WHO report, about 422 million people worldwide have diabetes, the majority living in low-and middle-income countries, and 1.5 million deaths are directly attributed to diabetes each year. Both the number of cases and the prevalence of diabetes have been steadily increasing over the past few decades ².

Post-prandial hyperglycemia still remains a problem in the management of type 2 diabetes mellitus. Of all available anti-diabetic drugs, α-glucosidase inhibitors seem to be one of the most effective in reducing post-prandial hyperglycemia ³.

Alpha-amylase and alpha-glucosidase are the key enzymes responsible for metabolism of carbohydrates. Alpha-glucosidase Inhibitors (AGIs) are oral anti-diabetic drugs preferably for treatment of T2DM. AGIs delay the process of carbohydrate absorption in the gastrointestinal tract by moving the undigested carbohydrate into the distal part of small intestine and colon. This class of drugs helps in reduction in postprandial hyperglycaemia ⁴. AGIs are saccharides that act as competitive inhibitors for the enzymes in the small intestine to slow down the digestion of carbohydrates such as starch, so that glucose from food enters the bloodstream more slowly, leading to the reduction in postprandial hyperglycaemia (Fig.1). 

There are three FDA approved AGIs available in the market (Fig. 2).  Acarbose obtained from Actinomyces utahensis was the first AGIs, used as a competitive inhibitor of α-glucosidase ⁵. Voglibose and Miglitol are the other AGIs used for management of Type II diabetes ⁶. These drugs have benefits in reducing post-prandial sugars when usually combined with other anti-diabetic drugs and thus lower HbA1c ⁷. These facilitate to raise post meal levels of GLP-1 that subsequently delays digestion and decreases appetite ⁸.

 

 

 

Figure 1: Mechanism of α-glucosidase inhibitors for controlling post-prandial glucose levels

 

 

Figure 2: FDA approved α-glucosidase inhibitors

 

 

Side effects of AGIs typically include bloating, flatulence, gastrointestinal irritation that might be recovered in few weeks ⁹. α-glucosidase inhibitors are not recommended if the person has any kind of GIT related disorder like ulcerative colitis or Crohn’s disease, blockage in intestines, digestive disorder in intestines, diabetic ketoacidosis; a condition where body burns fat instead of carbohydrates for energy ⁹. Acarbose is not recommended if the patient has an ulcer in large intestine, cirrhosis of the liver and for pregnant women ^{9, 10}

On the basis of these literature observations, it was thought worthwhile to identify some new α-glucosidase inhibitors with better safety profile therefore drug repurposing approach in combination with QSAR was considered to be better choice.

Drug repurposing is gaining popularity as a quick and effective method of identifying new therapeutic indications of approved drugs unrelated to their original medical intent, and is successfully moving towards the second phase of clinical trials. In this study, drug repurposing with QSAR based virtual screening was implemented for identification of some α-glucosidase inhibitors as new anti-diabetics.

 To carried of QSAR modeling against this target, congeneric series of N’-Benzylidenebenzoylhydrazide ^11-13 having two different types of substitutions on Benzylidene and benzoyl part as well as proper variation in the biological activity was selected on the basis the of thumb rules described by Hansch in his manual ¹⁴.

2. MATERIALS AND METHOD

QSAR analysis was carried out for α-glucosidase inhibitors using various conventional QSAR approaches including Free Wilson, Hansch, and Mixed modeling. For this purpose, various QSAR descriptors were collected from different sources like Hansch Manual, Medicinal Chemistry books etc.^14,15 Indicator variables for deriving Free Wilson approach were formulated from the various substituents present on the parent scaffold. Global properties of the molecules were calculated by PaDEL software ¹⁶. QSAR models were derived by DTC QSAR modeling tool ¹⁷. Internal and external validations were carried out by calculating various statistical parameters lilke Q², R²_traing, R² test PRESS values etc.  Virtual screening was carried out using DRUGBANK data base ¹⁸.  

3. RESULTS AND DISCUSSION  

For QSAR modeling, data set of N’-Benzylidenebenzoylhydrazide scaffold based α-glucosidase inhibitors containing 49 molecules was divided into training set of 35 molecules and test set of 14 molecules. Details about training set and test set are given in the Table 1. Training set was used for determining internal predictive ability whereas test set was used for external predictive ability of the QSAR model. Inhibitory activity data i.e. IC₅₀ was collected from the literature. IC₅₀ of the compounds represent their doses in micromolar concentration (μM) required to produce 50% inhibition of α-glucosidase enzyme. The given IC₅₀ data is first converted into pIC₅₀ by taking negative log of IC_50, where IC₅₀ is in molar concentration. The values of pIC₅₀ of all molecules in the data set are described in Table 1. 

3.1 Free-Wilson model

QSAR modeling was started with Free Wilson approach.   For this purpose various indicator variables were recorded for different functionality at R1 by assigning value 1 for presence of the particular group and value 0 for absence of that group. Indicator values are described in Table 2. Free Wilson model was developed by taking pIC₅₀ as dependent variable and various indicator variables as independent variables. The best model generated is given in Equation 1.   Evaluation parameters of this model clearly indicate that Model Quality is poor and therefore this model cannot be accepted for further analysis.  

pIC₅₀=4.64304(+/-0.07845)-1.45673(+/-0.31379) R3-OCH3 -1.41197(+/-0.43677) R2-Br -1.33974(+/-0.43677) R3-CH3 -1.20817(+/-0.43677) R4-CH3.....................................................1

Descriptions about selected variables are as follows: 

R3-OCH3 => 'Negative Contribution' 

R2-Br => 'Negative Contribution' 

R3-CH3 => 'Negative Contribution' 

R4-CH3 => 'Negative Contribution'

 

 

 

Table 1: Training set and test set data for QSAR analysis of α-glucosidase inhibitors

 

          * Test set compounds,  ^a Dose in micromolar concentration required to produce 50%     

              inhibition of  α-glucosidase^{,,  b} –log IC₅₀

Total number of compounds: 49

Number of trainings: 35, number of tests: 14

     

Internal Validation Parameters:

SEE :0.4296, r^2 :0.5951, r^2 adjusted :0.541, PRESS :5.5384, F :11.0264 (DF :4, 30)

Leave-One-Out(LOO) Result :

Q2 :0.2954

Rm^2metrics (after scaling the data): Average rm^2(LOO):0.15569, Delta rm^2(LOO):0.28955

External Validation Parameters (Without Scaling):

r^2 :0.03782, r0^2 :-1.3350, reverse r0^2:-1.4060, RMSEP:0.8305, Q2f1/R^2(Pred) :-1.1619, Q2f2 :-1.37693

External Validation Parameters (After Scaling):

Average rm^2(test) :-0.00375, Delta rm^2(test) :0.00595

Error Based Judgement of Test Set Predictions:

Mean Absolute Error (MAE; 95% data): 0.56031

Standard Deviation of Absolute Error (SD; 95% data): 0.4517

Model Quality based on MAE-based criteria: 'BAD'

Golbraikh and Tropsha acceptable model criteria’s[19]:

1. Q^2 0.2954   Failed* (Threshold value Q^2>0.5)

2. r^2 0.0378 Failed*  (Threshold value r^2>0.6)

3. |r0^2-r'0^2|0.0709            Passed   (Threshold value |r0^2-r'0^2|<0.3)

4. k 1.0247[ (r^2-r0^2)/r^2]   36.3040, OR*k'0.9452 [(r^2-r'0^2)/r^2]  8.1813   Failed*

Table 2: Indicator variables of Training set and test set molecules of α-glucosidase inhibitors for Free Wilson model 

S. N.	pIC50	R2-OH	R3-OH	R4-OH	R5-OH	R6-OH	R2-OCH3	R3-OCH3	R4-OCH3	R5-OCH3	R6-OCH3	R2-CH3	R3-CH3	R4-CH3	R2-Cl
1	4.812	1	0	1	0	1	0	0	0	0	0	0	0	0	0
2	4.772	1	1	0	0	0	0	0	0	0	0	0	0	0	0
3	4.564	1	0	1	0	0	0	0	0	0	0	0	0	0	0
4	4.423	1	0	0	1	0	0	0	0	0	0	0	0	0	0
5	4.764	0	1	1	0	0	0	0	0	0	0	0	0	0	0
6	4.427	1	0	0	0	0	0	0	0	0	0	0	0	0	0
7	4.064	0	1	0	0	0	0	0	0	0	0	0	0	0	0
8	4.562	0	0	1	0	0	0	0	0	0	0	0	0	0	0
9	4.532	1	0	0	0	0	0	0	1	0	0	0	0	0	0
10	4.463	0	1	0	0	0	0	0	1	0	0	0	0	0	0
11	4.427	1	0	0	0	0	0	0	0	1	0	0	0	0	0
12	3.205	0	0	0	0	0	0	1	0	1	0	0	0	0	0
13	3.231	0	0	1	0	0	0	0	0	0	0	0	0	0	0
14	3.312	0	0	0	0	0	0	0	0	0	0	1	0	0	0
15	3.303	0	0	0	0	0	0	0	0	0	0	0	1	0	0
16	3.435	0	0	0	0	0	0	0	0	0	0	0	0	1	0
17	4.529	0	0	0	0	0	0	0	0	0	0	0	0	0	1
18	4.190	0	0	0	0	0	0	0	0	0	0	0	0	0	0
19	4.417	0	0	0	0	0	0	0	0	0	0	0	0	0	0
20	3.909	0	0	0	0	0	0	0	0	0	0	0	0	0	0
21	4.009	0	0	0	0	0	0	0	0	0	0	0	0	0	0
22	4.054	0	0	0	0	0	0	0	0	0	0	0	0	0	0
23	4.767	0	0	0	0	0	0	0	0	0	0	0	0	0	0
24	4.642	0	0	0	0	0	0	0	0	0	0	0	0	0	0

 

Conti….

S.no	R3-Cl	R4-Cl	R2-NO2	R3-NO2	R4-NO2	R2-F	R3-F	R4-F	R1-H	R2-H	R3-H	R4-H	R5-H	R6-H	R2-Br	R3-Br
1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
2	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
3	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
4	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
5	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
6	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
7	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
8	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
9	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
10	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
11	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
12	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
13	0	0	0	0	0	0	0	0	0	0	0	0	0	0	1	0
14	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
15	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
16	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
17	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
18	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0	0
19	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0	0
20	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0	0
21	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0	0
22	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0	0
23	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0	0
24	0	0	0	0	0	0	1	0	0	0	0	0	0	0	0	0

 

Conti…….

S. N.	pIC50	R2-OH	R3-OH	R4-OH	R5-OH	R6-OH	R2-OCH3	R3-OCH3	R4-OCH3	R5-OCH3	R6-OCH3	R2-CH3	R3-CH3	R4-CH3	R2-Cl
25	4.712	0	0	0	0	0	0	0	0	0	0	0	0	0	0
26	3.167	0	0	0	0	0	0	1	0	0	0	0	0	0	0
27	3.161	0	0	0	0	0	0	0	1	0	0	0	0	0	0
28	5.187	1	0	0	0	0	0	0	0	1	0	0	0	0	0
29	4.947	0	0	1	0	0	0	0	0	0	0	0	0	0	0
30	4.950	1	1	0	0	0	0	0	0	0	0	0	0	0	0
31	5.157	1	0	0	0	0	0	0	1	0	0	0	0	0	0
32	5.256	1	0	0	0	0	0	0	0	0	0	0	0	0	0
33	4.821	0	1	1	0	0	0	0	0	0	0	0	0	0	0
34	5.253	1	0	1	0	0	0	0	0	0	0	0	0	0	0
35	4.579	0	0	0	0	0	0	0	0	0	0	0	0	0	0
36	5.148	0	0	0	0	0	0	0	0	0	0	0	0	0	0
37	4.791	0	0	0	0	0	0	1	0	1	0	0	0	0	0
38	5.094	0	0	0	0	0	0	0	0	0	0	0	0	0	0
39	4.553	0	0	0	0	0	0	0	0	0	0	0	0	0	0
40	4.737	0	0	0	0	0	0	0	1	0	0	0	0	0	0
41	5.077	1	0	1	0	1	0	0	0	0	0	0	0	0	0
42	5.093	0	0	1	0	0	0	0	0	0	0	0	0	0	0
43	5.275	0	0	0	0	0	0	0	0	0	0	0	0	0	0
44	4.955	0	0	0	0	0	0	1	0	0	0	0	0	0	0
45	5.036	0	1	0	0	0	0	0	0	0	0	0	0	0	0
46	4.273	0	0	1	0	0	0	0	0	0	0	0	0	0	0
47	4.349	0	1	0	0	0	0	0	1	0	0	0	0	0	0
48	4.926	1	0	0	1	0	0	0	0	0	0	0	0	0	0
49	4.954	0	1	0	1	0	0	0	0	0	0	0	0	0	0

 

Conti… 

 

Table 3: Substituent constants for Hansch Model

S. N.	pIC50	R2-Pi	R3-Pi	R4-Pi	R5-Pi	R6-Pi	Total-Pi	R2-MR	R3-MR	R4-MR	R5-MR	R6-MR	Total MR	R2-L	R3-L	R4-L	R5-L	R6-L	Total-L	R2-B1
1	4.812	-0.67	0	-0.67	0	-0.67	-1.34	0.29	0	0.29	0	0.29	0.58	2.74	0	2.74	0	2.74	8.22	1.35
2	4.772	-0.67	-0.67	0	0	0	-1.34	0.29	0.29	0	0	0	0.58	2.74	2.74	0	0	0	5.48	1.35
3	4.564	-0.67	0	-0.67	0	0	-1.34	0.29	0	0.29	0	0	0.58	2.74	0	2.74	0	0	5.48	1.35
4	4.423	-0.67	0	0	-0.67	0	-1.34	0.29	0	0	0.29	0	0.58	2.74	0	0	2.74	0	5.48	1.35
5	4.764	0	-0.67	-0.67	0	0	-1.34	0	0.29	0.29	0	0	0.58	0	2.74	2.74	0	0	5.48	0
6	4.427	-0.67	0	0	0	0	-0.67	0.29	0	0	0	0	0.29	2.74	0	0	0	0	2.74	1.35
7	4.064	0	-0.67	0	0	0	-0.67	0	0.29	0	0	0	0.29	0	2.74	0	0	0	2.74	0
8	4.562	0	0	-0.67		0	-0.67	0	0	0.29	0	0	0.29	0	0	2.74	0	0	2.74	0
9	4.532	-0.67	0	-0.02	0	0	-0.69	0.29	0	0.79	0	0	1.08	2.74	0	3.98	0	0	6.72	1.35
10	4.463	0	-0.67	-0.02	0	0	-0.69	0	0.29	0.79	0	0	1.08	0	2.74	3.98	0	0	6.72	0
11	4.427	0	0	-0.67	-0.02	0	-0.69	2.09	0	0	0.79	0	2.88	2.74	0	0	3.98	0	6.72	1.35
12	3.205	0	-0.02	0	-0.02	0	-0.04	0	0.79	0	0.79	0	1.58	0	3.98	0	3.98	0	7.96	0
13	3.231	0.86	0	-0.67	0	0	0.19	0.89	0	0.29	0	0	1.18	3.82	0	2.74	0	0	6.56	1.95
14	3.312	0.56	0	0	0	0	0.56	0.57	0	0	0	0	0.57	2.87	0	0	0	0	2.87	1.52
15	3.303	0	0.56	0	0	0	0.56	0	0.57	0	0	0	0.57	0	2.87	0	0	0	2.87	0
16	3.435	0	0	0.56	0	0	0.56	0	0	0.57	0	0	0.57	0	0	2.87	0	0	2.87	0
17	4.529	0.71	0	0	0	0	0.71	0.6	0	0	0	0	0.6	3.52	0	0	0	0	3.52	1.8
18	4.190	0	0.71	0	0	0	0.71	0	0.6	0	0	0	0.6	0	3.52	0	0	0	3.52	0
19	4.417	0	0	0.71	0	0	0.71	0	0	0.6	0	0	0.6	0	0	3.52	0	0	3.52	0
20	3.909	-0.28	0	0	0	0	-0.28	0.74	0	0	0	0	0.74	3.44	0	0	0	0	3.44	1.7
21	4.009	0	-0.28	0	0	0	-0.28	0	0.74	0	0	0	0.74	0	3.44	0	0	0	3.44	0
22	4.054	0	0	-0.28	0	0	-0.28	0	0	0.74	0	0	0.74	0	0	3.44	0	0	3.44	0
23	4.767	0.14	0	0	0	0	0.14	0.09	0	0	0	0	0.09	2.65	0	0	0	0	2.65	1.35
24	4.642	0	0.14	0	0	0	0.14	0	0.09	0	0	0	0.09	0	2.65	0	0	0	2.65	0

 

Conti…. 

 

Conti.. 

S.no	pIC50	R2-Pi	R3-Pi	R4-Pi	R5-Pi	R6-Pi	Total-Pi	R2-MR	R3-MR	R4-MR	R5-MR	R6-MR	Total MR	R2-L	R3-L	R4-L	R5-L	R6-L	Total-L	R2-B1
25	4.712	0	0	0.14	0	0	0.14	0	0	0.09	0	0	0.09	0	0	2.65	0	0	2.65	0
26	3.167	0	-0.02	0	0	0	-0.02	0	0.79	0	0	0	0.79	0	3.98	0	0	0	3.98	0
27	3.161	0	0	-0.02	0	0	-0.02	0	0	0.79	0	0	0.79	0	0	3.98	0	0	3.98	0
28	5.187	-0.67	0	0	-0.02	0	-0.69	0.29	0	0	0.79	0	1.08	2.74	0	0	3.98	0	6.72	1.35
29	4.947	0	0	-0.67	0	0	-0.67	0	0	0.29	0	0	0.29	0	0	2.74	0	0	2.74	0
30	4.950	-0.67	-0.67	0	0	0	-1.34	0.29	0.29	0	0	0	0.58	2.74	2.74	0	0	0	5.48	1.35
31	5.157	-0.67	0	-0.02	0	0	-0.69	0.29	0	0.79	0	0	1.08	2.74	0	3.98	0	0	6.72	1.35
32	5.256	-0.67	0	0	0	0	-0.67	0.29	0	0	0	0	0.29	2.74	0	0	0	0	2.74	1.35
33	4.821	0	-0.67	-0.67	0	0	-1.34	0	0.29	0.29	0	0	0.58	0	2.74	2.74	0	0	5.48	0
34	5.253	-0.67	0	-0.67	0	0	-1.34	0.29	0	0.29	0	0	0.58	2.74	0	2.74	0	0	5.48	0
35	4.579	0	0	-0.28	0	0	-0.28	0	0	0.74	0	0	0.74	0	0	3.44	0	0	3.44	0
36	5.148	0	0	0.14	0	0	0.14	0	0	0.09	0	0	0.09	0	0	0	2.65	0	2.65	0
37	4.791	0	-0.02	0	-0.02	0	-0.04	0	0.79	0	0.79	0	1.58	0	3.98	0	3.98	0	7.96	0
38	5.094	0	0	0	0	0	0	0.1	0.1	0.1	0.1	0.1	0.4	2.06	2.06	2.06	2.06	2.06	10.3	1
39	4.553	0	0.86	0.14	0	0	1	0	0.89	0.09	0	0	0.98	0	3.82	2.65	0	0	6.47	0
40	4.737	0	0	-0.02	0	0	-0.02	0	0	0.79	0	0	0.79	0	0	3.98	0	0	3.98	0
41	5.077	-0.67	0	-0.67	0	-0.67	-1.34	0.29	0	0.29	0	0.29	0.58	2.74	0	2.74	0	0	5.48	1.35
42	5.093	0	0.86	-0.67	0	0	0.19	0	0.89	0.29	0	0	1.18	0	3.82	2.74	0	0	6.56	0
43	5.275	0	0	0.71	0	0	0.71	0	0	0.6	0	0	0.6	0	0	3.52	0	0	3.52	0
44	4.955	0	-0.02	0	0	0	-0.02	0	0.79	0	0	0	0.79	0	3.98	0	0	0	3.98	0
45	5.036	0	-0.67	0	0	0	-0.67	0	0.29	0	0	0	0.29	0	2.74	0	0	0	2.74	0
46	4.273	0	-0.02	-0.67	0	0	-0.69	0	0.79	0.29	0	0	1.08	0	3.98	2.74	0	0	6.72	1.35
47	4.349	1.12	-0.67	-0.02	0	0	0.43	1.39	0.29	0.79	0	0	2.47	4.23	2.74	3.98	0	0	10.95	2.15
48	4.926	-0.67	-0.67	0	0	-0.67	-1.34	0.29	0	0	0.29	0	0.58	2.74	0	0	2.74	0	5.48	1.35
49	4.954	0	-0.67	0	-0.67	0	-1.34	0	0.29	0	0.29	0	0.58	0	2.74	0	2.74	0	5.48	0

 

Conti..

 

 

3.2 Hansch Model 

Data was further subjected to Hansch QSAR analysis by considering substituent’s constants. The values of various substituent’s constants were collected from Hansch QSAR manual ¹⁴ and given in the Table 3. Hansch model was developed by considering pIC₅₀ as dependent variable and substituent’s constants of R₁ as independent variables. No model was satisfied the acceptable statistical criterion. 

 

3.3 Mixed Model

Since no significant model was obtained from Hansch analysis, further attempt was made to develop mixed model by combining indicator variables and substituent’s constants in regression analysis. ‘Mixed Model’ was developed by considering pIC₅₀ as dependent variables and indicator variable and substituent’s constants simultaneously (given in Table 2 & 3) as independent variables. The best ‘Mixed model’ is given in Equation 2.     

pIC₅₀ = 4.69764(+/-0.09082)   -1.32531(+/-0.4518) R2-Br -1.51164(+/-0.31728) R3-OCH3 -1.39464(+/-0.43941) R3-CH3 -1.26264(+/-0.43941) R4-CH3 -0.50764(+/-0.43941) R3-Cl -0.15881(+/-0.17639) R2-MR………………………………………………………………………… 2

Description about selected variables are as follows: 

R2-Br => 'Negative Contribution' 

R3-OCH3 => 'Negative Contribution' 

R3-CH3 => 'Negative Contribution' 

R4-CH3 => 'Negative Contribution' 

R3-Cl => 'Negative Contribution' 

R2-MR => 'Negative Contribution' 

Internal Validation Parameters:

SEE :0.42992, r^2 :0.62175, r^2 adjusted :0.5407, PRESS :5.17522, F :7.67085 (DF :6, 28),Leave-One-Out(LOO) Result :, Q2 :0.39218, Rm^2 metrics (after scaling the data):

Average rm^2(LOO):0.19393, Delta rm^2(LOO):0.37955

External Validation Parameters(Without Scaling):

r^2 :0.0327, r0^2 :-1.3410, reverse r0^2:-1.3448, RMSEP:0.8288, Q2f1/R^2(Pred) :-1.1532, Q2f2 :-1.3674, 

External Validation Parameters (After Scaling):

Average rm^2(test):-0.00284, Delta rm^2(test):0.0051

ERROR-BASED JUDGEMENT OF TEST SET PREDICTIONS 

Mean Absolute Error (MAE; 95% data): 0.5518, 

Standard Deviation of Absolute Error (SD; 95% data): 0.4592, 

Model Quality based on MAE-based criteria: 'BAD'

Statistical parameters of Mixed Model (Equation 2) demonstrated BAD Model Quality. Therefore, Hansch type QSAR model development was further tried by considering some global properties of the molecules which were calculated by PaDEL software ¹⁶.  

3. 4 Hansch type Model using PaDEL descriptors

Data set was further subjected to Hansch type model development considering some global properties which were calculated by PaDEL software ¹⁶. The best model is described in Equation 3.

pIC₅₀ = 50.45765(+/-6.99755)   -0.23993(+/-0.11028) ETA_Alpha -0.35794(+/-0.05332) minHBa -21.25182(+/-3.70923) SpMin2_Bhs -2.33671(+/-1.53196) SpMin7_Bhm -1.10024(+/-0.98223) MATS4m -0.0223(+/-0.01685) minHBint8 …………………………………………. 3

ETA_Alpha(PaDEL; 2D)=> 'Negative Contribution' =>Sum of alpha values of all non-hydrogen vertices of a molecule

minHBa(PaDEL; 2D)=> 'Negative Contribution' =>Minimum E-States for (strong) Hydrogen Bond acceptors

SpMin2_Bhs => 'Negative Contribution' => smallest eigenvalue n. 2 of Burden matrix weighted by I-state

SpMin7_Bhm => 'Negative Contribution' => smallest eigenvalue n. 2 of Burden matrix weighted by mass 

MATS4m(Dragon; 2D autocorrelations)=> 'Negative Contribution' =>Moran autocorrelation of lag 4 weighted by mass 

minHBint8(PaDEL; 2D)=> 'Negative Contribution' =>Minimum E-State descriptors of strength for potential Hydrogen Bonds of path length 8

 

 

Table 4: Correlation matrix for QSAR model given in Equation 3

	ETA_Alpha	minHBa	SpMin2_Bhs	SpMin7_Bhm	MATS4m	minHBint8
ETA_Alpha	1.000
minHBa	02726	1.000
SpMin2_Bhs	-0.1801	-0.6902	1.000
SpMin7_Bhm	0.4003	0.2478	0.0036	1.000
MATS4m	0.2923	0.5778	-0.3145	0.1640	1.000
minHBint8	-0.0317	-0.0780	-0.1468	-0.2012	-0.1131	1.000

 

Internal Validation Parameters:

SEE :0.3020, r^2 :0.8133, r^2 adjusted :0.7733, PRESS :2.5543, F :20.3295 (DF :6, 28)

Leave-One-Out(LOO) Result :

Q2 :0.7466

Rm^2 metrics (after scaling the data):

Average rm^2(LOO):0.6604, Delta rm^2(LOO):0.0768

External Validation Parameters(Without Scaling):

r^2 :0.8050, r0^2 :0.7468, reverse r0^2:0.8046, RMSEP:0.2915, Q2f1 or R^2(Pred) :0.7336, Q2f2 :0.7071

External Validation Parameters (After Scaling):

Average rm^2(test):0.6565, Delta rm^2(test) :0.1754

ERROR BASED JUDGEMENT OF TEST SET PREDICTIONS:

Mean Absolute Error (MAE; 95% data): 0.2047, 

Standard Deviation of Absolute Error (SD; 95% data): 0.1074

Model Quality based on MAE-based criteria: 'MODERATE'

Golbraikh and Tropsha acceptable model criteria's[19] :

1. Q^                        0.7466 Passed   (Threshold value Q^2>0.5)

2. r^2            0.8050 Passed   (Threshold value r^2>0.6)

3. |r0^2-r'0^2| 0.0577 Passed   (Threshold value |r0^2-r'0^2|<0.3)

4. k 1 .0236   [(r^2-r0^2)/r^2]   0.0723 OR*   k' 0.9735    [(r^2-r'0^2)/r^2]   0.0005 Passed

(Threshold value: [0.85<k<1.15 and ((r^2-r0^2)/r^2)<0.1 ] OR* [0.85<k'<1.15 and ((r^2-r'0^2)/r^2)<0.1] )

 

 

Statistical evaluation of Equation 3 clearly demonstrated that model is having acceptable values of primary statistical parameters including SEE: 0.3020, r² : 0.8133, r² adjusted :0.7733, PRESS :2.5544, F :20.3295, Q² :0.7466 which determine internal consistency and  r² :0.8050, r₀² :0.7468, reverse r₀² : 0.8046, RMSEP:0.2915, Q²f1 or R2(Pred) :0.7336, Q2f2 :0.7071 Average rm^2(test) :0.6565, Delta rm^2(test) :0.1754 which determine external predictive ability. Other criterion including Model Quality based on MAE-based criteria and Golbraikh and Tropsha ⁵ acceptable model criteria's also pass the model for its acceptability to use it for screening α-glucosidase inhibitors and prediction of their activities. Predicted activities of training and test set molecules from the best model, Equation 3, along with residual values are given in Table 5. Graph of observed vs predicted activity from this model is shown in Fig. 2 and compound vs residual is shown in Fig. 3.  These graphs clearly indicate that most of the compounds predicted within ± 0.5 pIC₅₀ units. 

 

 

Table 5: Predicted activities of training and test set molecules along with residual values

* Test compounds, ^a - logIC_50, where IC₅₀ is experimental reported in the literature, 

^b predicted –log(IC₅₀) from the best model Equation 3, ^c OBEpIC₅₀-PRED.pIC₅₀

 

Figure 2: Graph of observed Vs. Predicted activities of training and test sets from Equation 3

Figure 3: Resisual plot for training and test set

 

 

Some α-glucosidase inhibitors were identified by QSAR model based virtual screening (VS) protocol. VS are a computational technique used in identification new bioactive molecules. It deals with the quick search of large libraries of chemical structures in order to identify those structures which are most likely to map over the query in silico model. For this purpose, the best QSAR model of α-glucosidase inhibitors, given in Equation 4.3, was used to screen out some α-glucosidase inhibitors as NCE with anti-diabetic effect. These best models were used as filters for screening DRUGBANK using Predict Module of DTC QSAR tool ^{17, 18}. To predict activities of the screened out molecules, descriptors of these were calculated by PaDEL software ¹⁶. Some identified α-glucosidase inhibitors along with their predicted in along with predicted pIC₅₀ from Equation 3 is given in Table 6. Top ten repurposed α-glucosidase inhibitors screened out by virtual screening using Equation 3 as query against DRUGBANK are shown in Fig.4.

 

 

Table 6: Newly identify α-glucosidase inhibitors as anti-diabetic drug

 

Figure 4: Top ten repurposed α-glucosidase inhibitors screened out by virtual screening using Equation 3 as query against DRUGBANK.

 

4. CONCLUSION

On the basis of this QSAR modeling it can conclude that a Hansch type 2D QSAR model has been successfully developed by utilizing some 2D PaDEL descriptors. Generated model was thoroughly evaluated by means of all reported statistical parameters. This validation results of the best model Equation 3 are in acceptable criterion and therefore suggest model’s reliability to be used in VS for identifying repurposed α-glucosidase inhibitors which may be further develop as new effective anti-diabetic in management of post-prandial hyperglycemia in the type II diabetes without additional safety measurement.

Acknowledgements: None 

Conflict of Interest: The authors declare no potential conflict of interest with respect to the contents, authorship, and/or publication of this article.

Author Contributions: All authors have equal contribution in the preparation of manuscript and compilation.

Source of Support: Nil

Funding: The authors declared that this study has received no financial support.

Informed Consent Statement: Not applicable. 

Data Availability Statement: The data presented in this study are available on request from the corresponding author. 

Ethical approvals: This study does not involve experiments on animals or human subjects.

REFERENCES

1. Pizzi RA. Defying diabetes: The discovery of insulin. Modern Drug Discovery 2000; 3(6):77-80.

2. APh A Special Report. New approaches to insulin therapy for diabetes. American Pharmaceutical Association, Washington DC 2001.

3. Derosa G, Maffioli P. α-Glucosidase inhibitors and their use in clinical practice, Arch Med Sci 2012; 5:899-906 https://doi.org/10.5114/aoms.2012.31621 PMid:23185202 PMCid:PMC3506243

4. Wehmeier U, Piepersberg W. Biotechnology and molecular biology of the α - glucosidase inhibitor acarbose, Appl. Microbiol. Biot. 2004; 63:613-625. https://doi.org/10.1007/s00253-003-1477-2 PMid:14669056

5. Narita T, Yokoyama H, Yamashita R, Sato T, Hosoba M, Morii T., et al., Comparisons of the effects of 12-week administration of miglitol and voglibose on the responses of plasma incretins after a mixed meal in Japanese type 2 diabetic patients, Diabetes. Obes. Metab. 2011; 14:283-287. https://doi.org/10.1111/j.1463-1326.2011.01526.x PMid:22051162

6. Derosa G, Mereu R, D'Angelo A, Salvadeo S, Ferrari I, Fogari E, et al., Effect of pioglitazone and acarbose on endothelial inflammation biomarkers during oral glucose tolerance test in diabetic patients treated with sulphonylureas and metformin, J. Clin. Pharm. Ther. 2010; 35:565-579. https://doi.org/10.1111/j.1365-2710.2009.01132.x PMid:20831680

7. Derosa G, Maffioli P. Mini-Special Issue paper Management of diabetic patients with hypoglycemic agents α-Glucosidase inhibitors and their use in clinical practice, Arch. Med. Sci. 2012; 5:899-906. https://doi.org/10.5114/aoms.2012.31621 PMid:23185202 PMCid:PMC3506243

8. Holt R, Lambert K. The use of oral hypoglycaemic agents in pregnancy, Diabet. Med. 2014; 31:282-291. https://doi.org/10.1111/dme.12376 PMid:24528229

9. Syahrul I. et al. Synthesis of novel flavone hydrazones: In-vitro evaluation of α-glucosidase inhibition, QSAR analysis and docking studies Eur. J. Med. Chem., 2015; 105:156-170. https://doi.org/10.1016/j.ejmech.2015.10.017 PMid:26491979

10. Muhammad T. et al. Synthesis of novel inhibitors of α-glucosidase based on the benzothiazole skeleton containing benzohydrazide moiety and their molecular docking studies, Eur. J. Med. Chem, 2015; 92:387-400. https://doi.org/10.1016/j.ejmech.2015.01.009 PMid:25585009

11. Farman A, et al. Hydrazinyl arylthiazole based pyridine scaffolds: Synthesis, structural characterization, in vitro α-glucosidase inhibitory activity, and in silico studies, Eur. J. Med. Chem., 2017; 138:255-272. https://doi.org/10.1016/j.ejmech.2017.06.041 PMid:28672278

12. Flynn GL. Substituent constants for correlation analysis in chemistry and biology. By Corwin Hansch and Albert Leo. Wiley, 605 Third Ave., New York, NY 10016. 1979.

13. Golbraikh A, Tropsha A., Beware of Q2, J Mol Graph Model, 2002; 20:269-76. https://doi.org/10.1016/S1093-3263(01)00123-1 PMid:11858635

14. Krzywinski M, Altman N. Classification and regression trees. Nat Methods. 2017; 14(8):757. doi: 10.1038/nmeth.4370. https://doi.org/10.1038/nmeth.4370

15. Costa VG, Pedreira CE. Recent advances in decision trees: an updated survey. Artif Intell Rev. 2023; 56:4765-4800. https://doi.org/10.1007/s10462-022-10275-5

16. Ahn S, Lee SE, Kim M. Random-forest model for drug-target interaction prediction via Kullback-Leibler divergence. J Chemin- 2022; 14:67. https://doi.org/10.1186/s13321-022-00644-1 PMid:36192818 PMCid:PMC9531514

17. Kapsiani S, Howlin BJ. Random forest classification for predicting lifespan-extending chemical compounds. Sci Rep. 2021;11 https://doi.org/10.1038/s41598-021-93070-6 PMid:34226569 PMCid:PMC8257600

18. Yu F, Wei C, Deng P, Peng T, Hu X. Deep exploration of random forest model boosts the interpretability of machine learning studies of complicated immune responses and lung burden of nanoparticles. Sci Adv. 2021; 7 https://doi.org/10.1126/sciadv.abf4130 PMid:34039604 PMCid:PMC8153727

19. Tharwat A. Parameter investigation of support vector machine classifier with kernel functions. Knowl Inf Syst. 2019; 61:1269-1302. https://doi.org/10.1007/s10115-019-01335-4