*3.2. Structure Based Virtual Screening and Predicted Binding Affinities*

Initially, the Molecular docking methodology was validated by redocking a co-crystal ligand with targeted protein using the same coordinates. RMSD values of less than 2 angstrom were obtained, which demonstrate the successful validation of the docking protocol and can be used to describe ligand poses with specificity and accuracy. Afterward, virtual screening was conducted with 1200 ligands library and Dabrafenib. The coordinates of co-crystal ligand were used to dock the ligand library and Dabrafenib. Out of

1200 drug candidates, only four drug candidates showed excellent binding energies that were even better than Dabrafenib. The binding energies of top hits and Dabrafenib are tabulated in Table 4. In particular, compound **762** showed a maximum binding energy of −42.67 kJ/mol and exhibited strong binding affinity with NEK7 protein when subjected to the DL prediction model.


**Table 4.** Binding energies and Predicted binding affinities via Deep learning model.

Among the four top hits, compound **208** exhibited promising hydrophobic and hydrophilic interactions. The amino acid residues involved in important molecular interactions were as follows: ASP115, ARG121, GLY117, ASP179, PHE168, ILE195, ALA114, ALA61, ILE40, and ASP118. It was observed that two hydrogen bonds were involved in stabilizing the protein–ligand complex. One hydrogen bond was observed with ASP118 with a bond length of 2.97 angstrom, while the second hydrogen bond was observed with GLY117 having a bond length of 3.14 angstrom. Important residues of the active site were engaged in hydrophobic interactions, including van der Waals interactions, pi-alkyl and alkyl–alkyl interactions. The docking score of compound **208** was found to be −33.47 kJ/mol. Similarly, compound **248** exhibited stronger molecular interactions with the following amino acid residues: ARG121, ASP118, ALA165, LYS63, ASN166, ASP179, PHE168, LEU111, VAL48, ALA116, ASP115, and GLY117. It was discovered that important amino acid residues of the NEK7 protein's DLG/DFG motifs were involved in interactions. Furthermore, the amino acids LEU111 and LYS163 interacted via hydrophobic bonds. Two important hydrogen bonds were contributing toward the stability of conformations. One hydrogen bond engaged GLY117 residues with a bond length of 2.2 angstroms. Another hydrogen bond was engaging ASN166 amino acid with a bond length of 3.34 angstroms. Among hydrophobic interactions, van der Waals interactions played a pivotal role in stabilizing the complex. The docking score of the compound **248** was −35.56 kJ/mol. The putative 2D and 3D binding modes of compounds **208** and **248** are shown in Figure 5.

Another important top hit was compound **255**, which exhibited potent molecular interactions with amino acid residues of the active site. It was the second-best drug candidate obtained via virtual screening. Amino acid residues involved in bonding and nonbonding interactions were as follows: PHE45, SER46, LYS63, ALA114, ASP115, GLU112, PHE168, ASP179, VAL48, GLY43, and GLN44. It was observed that two important hydrogen bonds with short bond lengths were contributing toward stabilizing the complex. One hydrogen bond occurs between the electronegative oxygen atom of the compound **255** and the SER46 residue of the targeted protein. Moreover, the second hydrogen bond was engaged in GLY117 with a bond length of 3.16 angstroms. As shown in Figure 5, amino acid residues from the active site were involved in hydrophobic interactions with compound **255**.

**Figure 5.** The putative 2D and 3D binding mode of compound **208** (**A**) and **248** (**B**). Green dashes are indicating hydrogen bonding whereas red dashes are indicating hydrophobic interactions.

Now referring to the top hit obtained through SBVS, namely compound **762**, It has shown excellent docking scores and demonstrated significant binding affinity obtained through deep learning models. It was observed that compound **762** was engaged in three hydrogen bonds of moderate-to-strong strength. One hydrogen bond occurred between the pentazole ring of the compound **762** and the electronegative oxygen atom of TYR201. The bond length of interaction was 3.08 angstroms. Similarly, the second hydrogen bond engaged SER234 residues with a surprisingly smaller bond length of 2.92 angstroms. These interactions lend enough testimony to stronger molecular interactions and more stabilized protein–ligand complexes. Furthermore, the third and last hydrogen bond occurred between TYR237 and compound **762** with a bond length of 3.02 angstroms. All three amino acid residues involved in hydrogen bonding belong to the activation loop of the NEK7 protein. The remaining active site residues, ILE123, GLU228, PHE236, MET203, PRO200, LEU246 and LEU232, engaged in hydrophobic interactions with compound **762**. The docking score and binding affinity (IC50) were predicted to be best among all top hits, i.e., −42.67 kJ/mol and 61.74 nM, respectively. Compound **762** could be a promising drug candidate for the treatment of NEK7-associated malignancies. The binding interactions of compounds **255** and **762** are shown in Figure 6.

**Figure 6.** The putative 2D and 3D binding mode of compound **255** (**A**) and **762** (**B**). Green dashes are indicating hydrogen bonding whereas red dashes are indicating hydrophobic interactions.

The bonding and non-bonding interactions of standard Dabrafenib was involving important amino acid residues of NEK7 activation loop. ARG50, LYS38, ALA165, ILE40, GLY117, ASP115, PHE168, LEU111, LEU112, ALA114, LEU113, ALA161, ASP179, and ILE95 were the amino acid residues implicated in molecular interactions with Dabrafenib. Dabrafenib exhibited significant molecular interactions, which contributed towards complex binding affinity. The strong interactions were observed with targeted protein and sulphonamide rings. The sulphonamide ring was implicated in several important stabilizing contacts, including conventional hydrogen bonding with ASP115 of the activation loop, Pi-cation interaction with ARG50, and interactions with ILE40 and ASP115 by two fluorine atoms connected to the ring. PHE168 formed pi-cation and pi-pi T-shaped contacts with the butylthiazole ring, whereas the pyrimidine ring produced conventional hydrogen bonds with GLU112 and ASP179, a carbon–hydrogen connection with ALA114, and a pi-alkyl interaction with ALA161. Due to important chemical interactions, Dabrafenib has a good binding energy of −33.89 kJ/mol. van der Walls interactions are essential hydrophobic interactions that have been observed with the amino acids LYS38, ALA165, GLY117, LEU111, LEU113, and ILE95. Figure 7 depicts the probable binding mode of Dabrafenib with NEK7.

**Figure 7.** 2D and 3D interactions of NEK7–Dabrafenib complex.
