*Cheminformatics and Its Applications*

*Molecular Electrostatic Potential and Chemometric Techniques as Tools to Design Bioactive… DOI: http://dx.doi.org/10.5772/intechopen.89113*

#### **Figure 4.**

*Cheminformatics and Its Applications*

**56**

C C1 C2 3 O C1 4C3 C C3 N4 1 O C1 N4 1 C N4 O1 2 C N4 O1 3

O

**Table 1.**

*Experimental and theoretical structural parameter of the 5-nirofuran-2-aldoxime.*

N2 O1 3

127.3 *\*Refers to the base sets cited in the corresponding references.*

127.9

125.3

126.2

127.7

128.1

118.8

126.2

125.2

126.4

127.6

126.4

119.6

120.1

118.8

117.7

117.2

118.9

118.1

117.6

117.2

115.6

118.2

118.9

118.1

117.2

119.2

117.3

117.5

116.3

115.0

114.9

115.8

115.6

114.6

114.6

115.9

115.5

115.7

115.4

115.0

116.0

117.4

117.8

116.9

118.4

117.8

118.1

117.6

118.3

117.7

117.9

117.5

117.8

118.2

117.5

118.4

131.4

131.4

129.8

130.2

129.9

130.6

130.2

130.1

130.0

130.9

130.5

130.9

130.3

129.5

130.4

111.1

110.6

113.2

111.5

112.3

111.2

112.0

111.5

112.2

111.2

111.9

111.1

111.4

112.8

111.1

105.9

106.0

105.1

107.5

106.6

107.5

106.6

107.7

106.8

107.7

106.9

107.8

106.9

106.0

106.9

104.2

101.6

106

*Score plots of the two first PCs, PC1 and PC2, for the separation of the approaches basis sets into classes: semiempirical and semiempirical not.*

#### **Figure 5.**

*Dendrogram obtained with HCA technique for the separation of the approach basis set into two classes: semiempirical and semiempirical not.*

maps show negative regions ranging from −82.99 to −4.87 kcal/mol. In the most active compound (**6**), as can be seen, the most negative values are in the nitro group, the O atom of the furan ring and the O atoms of the ester group (red and yellow). Also, the MEP maps of these compounds exhibit positive regions between the +4.54 and + 76.96 kcal/ mol values (green and blue). Compounds with double unsaturation, containing N atom next to the carbonyl, raise the electronic density with the increase of the carbonic chain. In the most active compound (**7**), the MEP map shows a region of negative values between −77.74 and − 1.31 kcal/mol, with the electron density concentrating mainly on the atoms of the nitro group, on the O atom of the furanic ring and on the N and O atoms of the amide group (red and yellow). According to the MEP map, these compounds present positive MEP between +5.64 and 61.21 kcal/mol (green and blue).

(ii) Compounds with double unsaturation, containing O atom neighboring the carbonyl, raising the carbon chain, increase the electron density in the atoms of the

**Figure 6.** *MEP maps (kcal/mol) for nitrofurans (training set).*

nitro group, extending through the O atom of the furan ring to the O atoms of the ester group following the unsaturated chain. In these compounds (**10–12**), the MEP maps exhibit more negative values between −76.18 and − 6.36 kcal/mol (red and yellow). They exhibit positive MEP in the range of +0.63 to 67.42 kcal/mol (green and blue)

(iii) Compound with an unsaturation, N atom neighboring the carbonyl in the carbonic chain and bulky substituents, has higher electron density in the vicinity of the furan ring and in the N and O atoms of the amide group. In this compound

**59**

figure.

*Molecular Electrostatic Potential and Chemometric Techniques as Tools to Design Bioactive…*

(**23**), the MEP map shows a negative region (red and yellow) between −73.10 and − 1.59 kcal/mol on the mentioned atoms and positive region between +5.56 and 69.91 kcal/mol (green and blue). The electron density around the nitro group, the O atom of the furan ring, and other atoms may induce the nitrofurans to show antitrypanosomal activity, suggesting the complexation in those regions with the active

From the above discussion, as a rule, to plan more active nitrofurans, we can assume we resort to one of the basic structures of the most active compounds and introduce groups of atoms or substituents electron donors enhancing the key

To perform the chemometric modeling, all variables were auto-scaled as preprocessing so that they could be standardized and so they could have the same importance regarding the scale. Furthermore, given a large quantity of multivariate data available, it was necessary to reduce the number of variables. Thus if any two descriptors had a high Pearson correlation coefficient (r ˃ 0.8), one of the two was excluded from the matrix at random, since theoretically they describe the same property [88]; they also have a high correlation with antitrypanosomal activity, and only one of them is enough to be used as independent variable in a predictive model.

Four molecular descriptors were selected for PCA model. The molecular descrip-

tors (QN1, gap energy, Mor05u, and MlogP), in vitro *T. cruzi* growth inhibition (experimental data), and activity and correlation matrix including all data for 23 nitrofurans can be seen in **Table 2**. The correlation between descriptors is less than 0.786. The first three principal components (PCs) describing 96.48 of the original information for the 23 are as follows: 45.70, 30.91, and 19.87%. PC1-PC2 scores for the samples are shown in **Figure 7**. From this figure, we can see that the nitrofurans are distributed into two distinct regions in PC1. The more active compounds are on the left side (**4–7, 10–12, 18,** and **23**) and the less active on the right side (**1–3, 8, 9, 13–17,** and **19–22**). According to **Figure 8**, the MlogP descriptor is responsible for displaying more active compounds on the left side, while the gap energy, QN1, and Mor05u descriptors displayed fewer active compounds for the right side from this

**Table 3** shows the loading vectors for PC1, PC2, and PC3. According to this

PC1 = 0.20 (QN1) + 0.06 (Gap energy) + 0.71 (Mor05u) − 68 (MlogP). (2)

From this equation, more active nitrofurans, in general, can be obtained when we have lower values for the QN1 combined with lower values for Gap energy and

The results of the HCA model are displayed in the dendrogram in **Figure 9** and are similar to those of PCA model. The nitrofurans are fairly well grouped according to their activity. From this figure, the two clusters (+ and −) mirror the same two

table, PC1 can be expressed through the following equation:

Mor05u and higher values for MlogP.

classes displayed by PCA model (**Figure 7**).

*3.2.3.2 HCA model*

*DOI: http://dx.doi.org/10.5772/intechopen.89113*

*3.2.3 Chemometric modeling*

*3.2.3.1 PCA model*

site of the receptor in a biological recognition process.

structural features that are necessary for their activities.

#### *Molecular Electrostatic Potential and Chemometric Techniques as Tools to Design Bioactive… DOI: http://dx.doi.org/10.5772/intechopen.89113*

(**23**), the MEP map shows a negative region (red and yellow) between −73.10 and − 1.59 kcal/mol on the mentioned atoms and positive region between +5.56 and 69.91 kcal/mol (green and blue). The electron density around the nitro group, the O atom of the furan ring, and other atoms may induce the nitrofurans to show antitrypanosomal activity, suggesting the complexation in those regions with the active site of the receptor in a biological recognition process.

From the above discussion, as a rule, to plan more active nitrofurans, we can assume we resort to one of the basic structures of the most active compounds and introduce groups of atoms or substituents electron donors enhancing the key structural features that are necessary for their activities.

#### *3.2.3 Chemometric modeling*

*Cheminformatics and Its Applications*

**58**

(green and blue)

*MEP maps (kcal/mol) for nitrofurans (training set).*

**Figure 6.**

nitro group, extending through the O atom of the furan ring to the O atoms of the ester group following the unsaturated chain. In these compounds (**10–12**), the MEP maps exhibit more negative values between −76.18 and − 6.36 kcal/mol (red and yellow). They exhibit positive MEP in the range of +0.63 to 67.42 kcal/mol

(iii) Compound with an unsaturation, N atom neighboring the carbonyl in the carbonic chain and bulky substituents, has higher electron density in the vicinity of the furan ring and in the N and O atoms of the amide group. In this compound

To perform the chemometric modeling, all variables were auto-scaled as preprocessing so that they could be standardized and so they could have the same importance regarding the scale. Furthermore, given a large quantity of multivariate data available, it was necessary to reduce the number of variables. Thus if any two descriptors had a high Pearson correlation coefficient (r ˃ 0.8), one of the two was excluded from the matrix at random, since theoretically they describe the same property [88]; they also have a high correlation with antitrypanosomal activity, and only one of them is enough to be used as independent variable in a predictive model.

#### *3.2.3.1 PCA model*

Four molecular descriptors were selected for PCA model. The molecular descriptors (QN1, gap energy, Mor05u, and MlogP), in vitro *T. cruzi* growth inhibition (experimental data), and activity and correlation matrix including all data for 23 nitrofurans can be seen in **Table 2**. The correlation between descriptors is less than 0.786. The first three principal components (PCs) describing 96.48 of the original information for the 23 are as follows: 45.70, 30.91, and 19.87%. PC1-PC2 scores for the samples are shown in **Figure 7**. From this figure, we can see that the nitrofurans are distributed into two distinct regions in PC1. The more active compounds are on the left side (**4–7, 10–12, 18,** and **23**) and the less active on the right side (**1–3, 8, 9, 13–17,** and **19–22**). According to **Figure 8**, the MlogP descriptor is responsible for displaying more active compounds on the left side, while the gap energy, QN1, and Mor05u descriptors displayed fewer active compounds for the right side from this figure.

**Table 3** shows the loading vectors for PC1, PC2, and PC3. According to this table, PC1 can be expressed through the following equation:
