3.9. MM-GBSA – free binding energy calculation
A prime module of Schrodinger Suite, MM-GBSA was selected to calculate the relative binding free energy for the improvement of the docking score of final IFD hits. It utilizes a surface generalized Born model for a more significant demonstration of a solvent surface area. The relative binding energy represents by ΔGbind, with the following equation (Lyne et al., 2006)[112–120].
ΔGbind = ΔE + ΔGsolv + ΔGSA
In which,
ΔGbind= Binding Free Energy,
ΔE = Difference of energy minimization between receptor-ligand complex & the energies of receptor and ligand
Where, ΔE = Ecomplex − Ereceptor – Eligand,
ΔGsolv = Difference of electrostatic solvation energy of the receptor-ligand complex& the energies of receptor and ligand
Where, ΔGsolv = Gsolv(complex) − Gsolv (receptor) – Gsolv (ligand),
ΔG SA = Difference of Surface area energies of the receptor-ligand complex& the energies of receptor and ligand
Where, ΔGSA = GSA(complex) − GSA (receptor) – GSA (ligand)
3.10.Molecular Dynamics Simulation
Groningen Machine for Chemicals Simulations (GROMACS) version 5.1.4 (http://www.gromacs.org/) was utilized in determining the structural dynamics of protein-ligand complexes (Jorgensen et al., 1996; Adcock et al., 2006). Top complexes of the screen along with reference, were taken for the MDS analysis. All the ligands’ topologies were generated using a webserver – PRODRG (Schüttelkopf& Van Aalten, 2004). The GROMOS force field is applied for performing MDS. Explicit water molecules were defined employing SPC (Simple Point Charge) model and a cubic period box with 1.0 nm distance (minimum) were set between protein and edge of the box. Following these, the protonation states of amino acid residues were set as per pH 7.0and the system was neutralized by adding counter ions [121–131].
Consequently, the energy minimization was performed for all protein-ligand complexes using the steepest descent energy approach (1,000 ps). Later on, the whole system was equilibrated by executing a position-restrained dynamics simulation (NVT and NPT) at 300 K for 300 ps. Finally, the equilibrated systems were subjected to MDStoanalyse the dynamics stability with 50 ns at a constant temperature of 300 K, the pressure of 1 atm, and an integration time step of 2 femtoseconds. Herein, other parameters like isothermal and isobaric coupling constants were set at 0.1ps and 2ps, respectively, within the minimum distance between box edges and any protein atom equal to 2.0 nm. Using Origin pro8 (Essmann et al., 1995) program, the statistical analysis like RMSD, RMSF, RGYR, and Hydrogen bond for each complex were analysed and respective graphs were plotted [132–140].
3.11MM-PBSA – free binding energy calculation
The calculation of binding free energy of protein-ligand was carried out by using the MM-PBSA approach with the help of thegmmpbsa module in three steps. The potential energy in a vacuum is calculated in the previous step. Otherenergies (polar and non-polar solvation) were predicted in respective steps. Herein, the non-polar solvation energy was generated using the solvent-accessible surface area (SASA) model[141–148].
The free energy of binding is coming from the following theory:
ΔGbinding = G-complex – (G-protein + G-ligand)
Where,G-complex = Total free energy of the protein-ligand complex
G-protein &G-ligand = Total free energies of the separated form of protein and ligand in the solvent, respectively.
3.12Principal component analysis:
Principal component analysis (PCA) or essential dynamics (ED) is one of the most important approaches to reveal the dynamic nature of proteins. It is a specific method to explain the functionally relevant motions of protein by the combination of local fluctuations and collective motions. The protocol was applied to build the covariance matrix with the extraction of concerted motion from all trajectories using backbone with g-covar and g-anaeig modules of GROMACS on selected screened compounds as well as reference inhibitor. In this process, the diagonalization of the covariance matrix generates a set of eigenvectors. A specific eigenvalue explains the energetic impact of the component on the motion [149–158].
3.13 Boiled Egg-Plot:
In the present investigation, we have used Boiled EGG-Plot to predict gastrointestinal absorption and brain penetration [16][61]. Aside from distinction to efficacy and toxicity, many new drug development failures are responsible for indigent pharmacokinetics and bioavailability. Gastrointestinal (GI) absorption and blood-brain access are two pharmacokinetics behaviors that estimate various drug development mechanisms. The Brain orIntestinal Estimated permeation method (BOILED-Egg) is expected as a factual predictive model that works by computing two parameters, i.e., the lipophilicity and polarity of the molecules. Contemporary predictions for both brain and intestinal permeation are accessed from the two same physicochemical descriptors and impartially rendered into the molecular design, owing to the speed, accuracy, conceptual simplicity and clear graphical output of the BOILED-Egg plot model. It also contains several parameters such as MW, TPSA, MLOGP, GI, and BBB, to revamp the BOILED-Egg plot. It can be enforced in various frameworks, from the filtering of chemical libraries at the initial steps of drug discovery and development to drug candidates' evaluation for further development.
3.14 Blood-Brain Barrier (BBB), HIA, Toxicity and LD50 comparison
The physiochemical properties of the top 4 compounds such as ZINC13377936, ZINC14726791, ZINC35753, ZINC90797260 were analyzed using SwissADME server, and the properties such as Blood-Brain Barrier (BBB), HIA, CYP inhibition, AMES toxicity carcinogenicity, LD50 in rat etc. were used for a comparative study using R programming. Abar plot was created using the ggplot library in R programming to compare the above properties [16][61].