Only ligands that meet particular criteria, such as being in the correct tautomer and ionization states and having the correct bond ordering, can be used in virtual screening (Madhavi Sastry et al., 2013). The Gaussian version 09 approach and realistic density functional theory (DFT) methodologies were employed to accomplish substantial atomic enlargement (Hosen et al., 2021). We employed a modified version of the Gaussian code with modifications to its polarization capability premise set (DNP), B3LYP, and Gaussian version 09 capabilities to obtain the highest achievable accuracy (Mohapatra et al., 2021). The files that represent electron negativity, electron partiality, energy gap, synthetic potential, hardness, delicate quality, and electrophilicity were solved using the criteria listed in the order (Equations 1–8). Subatomic limit orbital charts (HOMO and LUMO) were then calculate using mathematical procedures of given equations. The modified particle was then stored in a PDB file. E gap =   E LUMO − E HOMO (1) I = − E LUMO (2) A = − E HOMO (3) χ = I + A 2 (4) ω = μ 2 2 η (5) µ = − I + A 2 (6) η = I − A 2 (7) S = 1 η (8)

The PASS Online resource is freely available on the internet (http://www.way2drug.com/passonline). This tool seeks to predict the biological activity spectra of organic compounds for over 4000 different types of biological activity using their structural formulae (Poroikov et al., 2019). Over 95% of the time, it properly predicts the outcome. Researchers examined the structure-activity relationships in the training set to create the forecast. This set contains information on the structures and biological functions of approximately 300,000 chemical molecules. We examine the advantages and disadvantages of this strategy. There is information available on how to interpret the forecast’s findings (Druzhilovskiy et al., 2016). The PASS Online website has real-world applications that prioritize chemical synthesis and biological testing based on prediction findings. New pharmacological medications are being created, and PASS Online is expected to play a growing role as a multidisciplinary academic research center in this area (Siddikey et al., 2022).

Pharmacokinetics refers to the mathematical study of the ADME characteristics of a drug in relation to its dose and duration. The computational drug design and development process helps optimize a molecular candidate into a viable treatment by assessing pharmacokinetic features early on (Zhou et al., 2016). Therefore, the pharmacokinetic parameters of the selected drugs were determined using the Swiss-ADME server (http://www.swissadme.ch/index.php) (Daina et al., 2017). By utilizing the server, one can observe and anticipate the drug’s various pharmacokinetic and pharmacodynamics features.

A number of different proteins contribute to the development of breast cancer in females. Proteins with PDB IDs 3W32 were chosen as breast cancer susceptibility proteins after careful consideration of literature, methodologies, resolution, and organisms utilized for protein isolation. The choosing factors of the protein PBD ID 3w32 is depicted in the Supplementary Table S1. Crystal structures of these proteins were sourced from the Protein Data Bank (PDB) maintained by the RCSB (https://www.rcsb.org) (Burley et al., 2023). In order to remove water molecules and protein ligands, proteins were purified using the PYMOL software (version 2.4.1) before protein production. The website CASTp (http://sts.bioe.uic.edu/castp/index.html?2was) was utilized to gather the active site residues of the target proteins. Information on the proteins that were produced is shown in Table 1.

In protein-ligand interactions, binding sites can be found by looking for well-known pockets. The protein’s binding site was examined using BIOVIA Discovery Studio Visualizer v19.1 (BIOVIA) following a PDB search for the known and experimentally verified protein structure in complex with the ligand (PDB ID: 3w32). As shown in Figure 1, the binding site obtained from the complex structure was used in the receptor grid construction during the molecular docking using the PyRx virtual screening tool.

To find the best hit candidates against the needed protein, a molecular docking simulation was performed using the PyRx program (Dallakyan and Olson, 2015). PyRx is a free, free and open-source virtual screening implementation that also includes the docking wizards AutoDock 4 and Vina. It may search a big database of molecules for a specific macromolecule with a medical application. We used the AutoDock Vina wizard with Pyrex’s default settings to simulate molecular docking. To begin, we used Pyrex’s conjugate gradient approach to reduce the energy of the chosen ligands in the Merc molecular force field (mmff94 force field). The final phase of docking was to convert these energy to PDBQT format. Autodocking was performed with grid box dimensions of 58.5633 X, 53.9165 Y, and 66.1292 Z, with center X, Y, and Z values of 18.8128, 25.7483, and 14.2748, respectively. All of the ligands and proteins had their surfaces covered with a grid box.

The conformation of the ligand to the protein binding site is the most important factor in determining a potential active conformation, binding affinity, and strain discipline related to the binding mechanism. Structure optimization and least energy conformation methods based on the solution phase and the same gas-phase energy can be used to accomplish this kind of binding. Because of the metal ions, the ligand-protein combination does not conform to the expectations of classical molecular mechanics (MM) (Bronowska, 2011). The development of scoring functions that may describe electronic structure, electronic transitions, and system-specific charges in a molecular system reaction has recently been greatly aided by QM-based calculations. Eighty to ninety percent of modern quantum mechanics (QM) calculations employ DFT. For this study, DFT quantum mechanical computations on a subset of molecules were required. We started by finding compounds with the best possible bond lengths, angles, and dihedral angles. The DFT of the compounds was then calculated using Gaussian version 09 (Abdelrhman et al., 2021). The B3LYP functional set was utilized by.

DFT. The DFT needed the 6311G basis set to describe the molecule’s electronic wave functions.

The degree of stability of the chosen candidate compounds when bound to the active site cavity of the target protein was assessed by applying molecular dynamics (MD) simulations on the complex structure for a duration of 100 nanoseconds. This technique was done in order to assess whether the binding was secure. A molecular dynamics (MD) simulation of the complicated structure was performed in the Schrodinger research edition using the ‘Desmond v3.6 Program’ (Bharadwaj et al., 2021). The Linux operating system was used to conduct this simulation. This process has to be followed in order to ascertain whether thermodynamic stability is present in the receptor-ligand pair. Using a pre-set TIP3P water model and making sure the volume didn't change during the procedure solved the issue. By positioning an orthorhombic periodic box shape at a distance of 10 s on either side of the border, this was possible. It was found that the right ions, such as Na+ and Cl-, were selected and then randomly distributed throughout the solvated system in order to electrically neutralize it. The concentration of salt was 0.15 M. The Desmond module defined a procedure for reducing and relaxing the ligand and protein after they had come together to form the solvated system. The characteristics of the OPLS-2005 force field were used in this technique. The NPT ensemble was kept at a temperature of 300 K and an atmospheric pressure of one (1.01325 bar) by using the Nose-Hoover temperature coupling and isotropic scaling technique after fifty PS recording intervals with an energy of 1.2 were completed. The goal of doing this was to provide precise temperature coupling.

Schrodinger’s Maestro interface version 9.5 was utilized in order to render each and every image that was captured during the computational modelling simulation. The MD simulation was deemed to be of sufficient quality, and the Simulation Interaction Diagram (SID), which is a component of the Desmond module of the Schrodinger package, was utilized in order to conduct an analysis of the simulation event. According to this evaluation, the MD simulation is capable of meeting the requirements that are considered acceptable. On the basis of the trajectory output, the root mean square deviation (RMSD), root-mean-square fluctuation (RMSF), protein-ligand contacts (P-L contacts), and hydrogen-bond interactions were utilized in order to assess the stability of the complex structure. This investigation was conducted with the purpose of determining whether or not the intricate construction was capable of withstanding a variety of stresses without deteriorating in its shape.

The RMSD is a statistic used in molecular dynamics (MD) modelling to calculate the average distance an atom moves in comparison to a standard over a given time period (Fukutani et al., 2021). Distance is relative when compared to a starting point in time. Following the alignment of the relative mean square deviation (RMSD) of the protein-fit ligand atoms from each time frame, a comparison to the reference time, in this case 100 nanoseconds is performed. This comparison is based on the RMSD of the protein’s structural atoms, which include the Ca, backbone, sidechain, and heavy atoms. It would be evident whether or not the alignment was satisfactory as soon as the reference time arrived. The RMSD required for an MD simulation of length x time steps can be determined using the equation shown below (Equation 9). R M S D x = 1 N   ∑ i = 1 N r i ′   t x − r i t r e f 2 (9) here, N specifies the total number of selected atoms, t r e f denotes the reference time, and r' denotes the position of the selected ref atom in frame x. After superimposing the reference frame, Tx defines the recording intervals.

In order to characterize and keep track of the local conformational shift that takes place within a protein structure, the RMSF is primarily applied (Ding and Peng, 2019). An MD simulation of a protein can be constructed by using the equation (Equation 10), which asks for the number of residues and the RMSF value. R M S F i = 1 T   ∑ t = 1 T < r i ′   t − r i t r e f 2 > (10)

In this case, T mainly denotes the trajectory time, r' denotes the chosen atoms’ position in the reference frame as overlaid on frame i, t r e f denotes the reference or given time, and (<>) denotes the average of the square distance over residue b.

The amount of toxicity of a chemical substance can be measured by determining the extent to which it poses a risk to humans or animals, or by determining if it has the power to destroy an organ. Toxicity evaluation refers to both of these procedures. Prior to the initiation of a drug study, an investigation of the potential detrimental effects of chemical substances must be conducted. The conduct of a toxicity test is commonly recognized as one of the most important and critical components of the pharmaceutical production process. As a result, the web-based pkCSM server (Pires et al., 2015) was used to successfully complete the evaluation of the toxicity of the compounds chosen.

The diversity of natural phytochemicals, along with their intriguing biological roles, sets them apart from synthetic phytochemicals. It is difficult to draw clear functional and structure-activity relationships regarding the effects of phytochemicals on biological systems’ activity (Efferth and Koch, 2011). This is mostly due to the complex interactions that take place inside physiological systems as well as the high concentration of phytochemicals that have structural similarities. In addition, a significantly larger number of phytochemicals likely exist in nature, given the vast number already discovered. Technological advancements in synthesis, along with the development of more effective methods for isolation and analysis, have increased and help to identify the novel phytochemicals as lead compounds for the treatments of various diseases (Avidon et al., 1982). Figure 2 was generated with ChemDraw Ultra 12.0 and shows the chemical formulas and two-dimensional structures of top two ligands or phytochemicals and one FDA-approved medication (D1: Abemaciclib). The supplemental ST-2 contains an abundance of ligand-related information, and the 2D chemical structures of the selected ligands is depicted in Supplementary Figure S1.

In the discipline of computational chemistry, it is common practice to employ quantum mechanical methods to determine the thermodynamic, molecular orbital, and electrostatic properties of molecules. Each calculated derivative was strengthened and geometrically changed using the Gaussian 09 program to produce better results. Using DFT, we were able to boost the molecular orbital and thermal properties, allowing us to make predictions. This theory is compatible with both the hybrid model of B3LYP (Becke, 3-parameter, Lee-Yang-Parr) and the Gaussian version 09 polarization function basis set 6-311G (split-valence basis set) (Singh and Singh, 2017). Each chemical compound’s electrical energy, dipole moment, enthalpy, and free energy were calculated. Figure 3 depicts the optimal geometry and structure of the top most phytochemicals, and the all-optimized structure are shown in Supplementary Figure S2.

Optimized chemical structures of ligands are crucial in drug discovery, molecular docking, and computational chemistry, as they represent the most stable, energy-minimized conformations. By optimizing ligand structures, researchers ensure accurate interaction predictions with target biomolecules, enhancing binding affinity and specificity. This process helps in rational drug design, reducing experimental costs and time. Furthermore, optimized ligands improve the accuracy of pharmacokinetic and pharmacodynamic modeling, leading to more effective and safer therapeutic agents.

Chemical descriptors (HOMO and LUMO) and FMOs control the kinetic stability and chemical reactivity of molecules, respectively. Molecular orbitals that are least occupied are called LUMOs, and those that are most populated are called HOMOs. The electronic absorption of molecules releases one electron, which is caused by the HOMO state. The LUMO state accepts the electron simultaneously, and an energy gap forms as a result. These properties—kinetic stability, chemical reactivity, and atomic electrical transmission—are built upon this energy gap. The larger the energy difference, the more stable the molecule is when its HOMO and LUMO are far apart. The reason behind this is that the DFT method is used to calculate the energy gap. However, chemical stability is negatively affected by a small energy gap when the distance between a molecule’s HOMO and LUMO is small (Ahamed et al., 2023) (Kobir et al., 2023) (Yu et al., 2022). The small energy gap is the root cause of the chemical instability. In Figure 4, the color radish brown is used to represent the molecules’ positive node in both HOMO and LUMO situations, whereas the color deep green is used to represent the molecules’ negative node.

The table below covers several medications and phytochemicals, as well as information regarding their reactivity and frontier molecular orbitals (HOMO-LUMO). The sign A denotes HOMO energy, the letter I denotes LUMO energy, the letter μ represents chemical potential, the letter η denotes hardness, the letter σ denotes softness, the letter X denotes electronegativity, and the letter ω denotes electrophilicity. When we examine the energy gap between each of the fifteen molecules, we can find that Ligand No. 07 has the highest chemical stability at 11.3746 eV and the lowest at 7.49021 eV. This information can be found in Table 2. The Supplemental ST-3 has an extremely comprehensive computation.

The HOMO energy varies between −10.4799 eV (Ligand 7) and −8.54465 eV (Ligand 8), while the LUMO energy ranges from 0.89471 eV (Ligand 7) to −1.92793 eV (Ligand 10). The energy gap (E_gap) spans from 7.49021 eV (Ligand 10) to 11.3746 eV (Ligand 7), indicating significant variations in electronic stability and reactivity among the ligands. The chemical potential (μ) ranges from −5.77861 eV (Ligand 2) to −4.31233 eV (Ligand 9), with electronegativity (Χ) following the same trend. Hardness (η) varies from 3.74510 eV (Ligand 10) to 5.68732 eV (Ligand 7), while softness (σ) exhibits an inverse trend, emphasizing the ligands’ adaptability to electronic perturbations. Electrophilicity index (ω) values range from 2.01933 eV (Ligand 7) to 4.29672 eV (Ligand 10), indicating differences in their propensity to accept electrons. Notably, D-1 exhibits moderate values across all parameters, serving as a reference for comparative evaluation. These quantum chemical descriptors provide insights into the ligands’ stability, reactivity, and potential applications in coordination chemistry, catalysis, and molecular design.

An essential part of computer-aided drug design is determining the ligands’ molecular electrostatic potential (MEP), which yields a charge distribution map based on electron availability and scarcity. Additionally, it shows where the protein and ligand bind and how the charges are distributed in three-dimensional ligand structures. When applied to ligand surface analysis, MEP analysis can further help pinpoint where ligands are vulnerable to attack from electrophiles and nucleophiles (Lakshminarayanan et al., 2021; Guerrab et al., 2022). Utilizing quantum chemistry techniques, the MEP map was generated, and Gaussian functions were assessed with the use of Gaussian fundamental sets. In this study, blue represents positive charge, red negative charge, and green neutral charge. A lower negative charge than a larger positive charge is observed in all of the discovered ligands, as shown in the MEP map of top ligands in Figure 5. In Supplementary Figure S3 contains MEP map for all ligands.

The PASS prediction assessment displays the details of the antiviral, antibacterial, antifungal, antiparasitic, anticarcinogenic, anticancer (breast cancer), and inhibitory effects of drugs and ligands on breast cancer-resistant proteins. It is used to assess the potential biological activity against the targeted disease. According to our investigation, the Ligand No. 07 has low to moderate antiviral activity (Pa > 0.454), Ligand No. 11 and 14 have moderate to high antibacterial activity (Pa> 0.599), and Ligand No. 15 has low to significantly high antifungal activity (Pa > 0.678). To top it all off, most ligands have excellent anti-carcinogenic activity, meaning they won't cause cancer when taken orally. Ligands 05 (Pa > 0.649), 08 (Pa > 0.577), 11, and 14 (Pa > 0.502), as well as No (Pa> 0.526), exhibit strong anticancer properties, according to additional research. After reviewing the PASS prediction data for breast cancer-resistant protein has been selected, and we found that Ligand No. 05 has a very good Pa value (Pa > 0.516) as illustrate in Supplementary Table S4.

Pharmacokinetics (PK), an important parameter in medicine design, explains the time it takes for the body to absorb, distribute, metabolize, excrete, and contaminate a drug or foreign chemical after administration. This word is commonly used by pharmacists. As a result, PK promotes effective drug design (Lavé et al., 2016). Table 4 displays information for certain ligands and the medicine on drug likelihood, pharmacokinetics, Lipinski’s rule of five, and other topics. A ligand must meet ADME standards before it may be evaluated for drug candidate certification. The five Lipinski rules must be followed: a molecular weight ranging from 150 to 500 g/mol, a limited number of hydrogen bond donors, a number of rotatable bonds, a high bioavailability score, and a topological surface area value ranging from 20 Å² to 130 Å² (Singh et al., 2022; Al Hasib et al., 2022).

The Table 3 reveals that, with the exception of Ligands 11, 12, 13, 14, and 15, all of the ligands follow Lipinski’s rule of five. In terms of molecular weight, all of the ligands have the necessary molecular weights (150–500 g/mol), with the exception of drug D-1 (506.59 g/mol) and Ligand No. 15 (594.52 g/mol). As a result, with the exception of Ligands 04, 11, 14, and 15 (which have six rotatable bonds) and Ligand D-1 (17 rotatable bonds), no ligand can have more than three rotatable bonds. Except for Ligands 11 (12) and 12, the hydrogen bond acceptor can only receive a maximum of 10 hydrogen bonds from any of the ligands. Ligands are the numbers 12, 15, and N0 13. Except for ligands 7, 8, 9, 11, 12, 13, 14, and 15, no other ligand contributes more than five hydrogen bonds. Ligands 8, 10, 11, 12, 13, 14, and 15 do not match the drug development criteria since their topological surface areas exceed the range of 202 Å² to 130 Å². Apart from Ligands 11, 12, 13, 14, and 15, all of the other ligands had acceptable bioavailability values. After evaluating the facts offered previously, a conclusion can be reached. Several ligands have been removed from consideration as prospective pharmaceutical candidates; nevertheless, ligands 01, 02, 03, 05, and 06 remain on the list.

To identify possible breast cancer medication candidates, auto-docking was carried out with selected proteins susceptible to breast cancer (PDB ID: 3W32). Table 4 displays the molecular docking simulation data with binding affinities. In molecular docking, three chosen proteins have the highest binding affinities when interacting with Ligands Nos. 10, 11, and 12. However, Ligands 11 and 12 do not match the ADME criteria listed in Table 4. Ligands No. 08, 09, and 10 meet all drug property criteria in this study, having binding affinities of −8.5 kcal/mol, −8.4 kcal/mol, and −8.8 kcal/mol, respectively.

Protein-ligand interactions (PLIs) and protein-protein interactions (PPIs) play critical roles in identifying possible drug candidates for a target protein in structure-based drug design and drug discovery (Zhao and Bourne, 2022; Fu et al., 2018). This is why it is referred to as a critical component of the process. On the other hand, this part of the therapeutic goal is very important in and of itself. Because of the unique structural properties of protein interactions with ligands, it is today regarded as one of the most difficult areas of drug development. Bond distance research on the principal protease of breast cancer proteins have been conducted, with the primary focus of the studies being on the interaction of medicinal medicines with 3w32. Figure 6 depicts the key interactions that proteins and ligands have with amino acid residues.

Data on hydrogen bonds, hydrophobic bonds, and bond distances between amino acids are shown in Table 5. One of the most important factors in selecting a potential drug candidate is the bond distance, with a value between 3.1 Å to 3.55 Å indicating a weak link and a value between 2.5 Å to 3.1 Å indicating a strong binding, according to previous studies (da Cunha Xavier et al., 2024). Since the ADME screening and molecular docking did not include any other Ligands, we looked at how proteins with PDB ID 3W32 interacted with Ligands No. 01, to 15, and standard D1. From the, it shows that ligand 08 forms GLN791 (2.92 Å) and MET793 (2.20 Å) two strong hydrogen bonds, LEU718 (3.62 Å), VAL726 (4.70Å), VAL726 (3.99 Å), VAL726 (4.95 Å), LEU844 (4.69Å), ALA743 (4.28 Å), ALA743 (4.85 Å), LYS745 (4.53 Å) other bonds with protein 3W32. The ligand 09 forms ASP837 (2.55 Å), ASP855 (3.30 Å), AR841 (2.26 Å), and ARG841 (2.74 Å) strong bonds, and VAL726 (3.97 Å), ARG841 (4.83 Å), ALA743 (4.91 Å), LYS745 (5.20 Å), and LYS745 (4.51 Å) other bonds. On the other hand, ligand 10 forms MET793 (2.35Å) strong bond, and LEU718 (5.21Å), VAL726 (5.27Å), VAL726 (4.39Å), ALA743 (4.56Å), ALA743 (5.13Å), ALA743 (3.69 Å), ALA743 (4.99 Å), LEU844 (5.05 Å), VAL726 (3.85 Å), and LEU844 (3.50 Å) other bonds. Besides the standard D1 form ASP855 (2.65 Å), ASP855 (2.83 Å), ASP855 (2.58 Å), and THR854 (2.77 Å) strong bond with protein. For more details on the various bond classifications and types, see supplementary SF-4 and Supplementary Table S5.

MD simulation modelling, which is employed in computer-aided drug discovery, may study the protein-ligand complex’s stability and intermolecular interactions in real time. In a controlled situation, it can also detect conformational changes in complicated systems. A 100 ns molecular dynamics simulation was used to investigate the protein’s structural changes during its interaction with the chosen ligand. Multimolecular activity was measured on final images acquired from the appropriate 100 ns trajectories. It was examined the findings of a molecular dynamics (MD) simulation, and this comprised solvent accessible surface area (SASA), radius of gyration (Rg), root mean square fluctuation (RMSF), and root mean square deviation.

The average dislocation change of a chosen group of atoms over a certain period of time in relation to a reference time can be found using an RMSD value. A protein-ligand complex’s optimal RMSD shift falls between 1 Å and 3 Å, or 0.1 and 0.3 ns. When the RMSD value exceeds the permitted limit, the protein’s structure has undergone a significant alteration. To determine the RMSD value of the essential protein in association with the selected molecule and ligands 08, 09, and 10, we performed a 100 ns MD simulation. Figure 7 displays the RMSDs, or relative standard deviations, of several complexes. The following protein-ligand complexes are displayed: Apo-protein (red), protein-3w32 and ligand-08 (purple), protein-3w32 and ligand-10 (green), and protein-3w32 and ligand-D1 (standard) (blue). In Figure 7D, you can see all of these complexes unified. Prior to making comparisons, the red Apo protein Apo’s RMSD is shown. Figure 7A shows that the average RMSD value ranges from 1.5 Å to 2.6 Å when comparing the protein-ligand combination 3w32_08 (purple color) to Apo (red color). Figure 7B shows that between 1.7 Å and 2.8 Å, the protein-ligand combination 3w32_09 (green) compared to Apo (red). Comparing the protein-ligand combination 3w32_10 (orange color) with Apo (red color), Figure 7C shows the RMSD variation within 1.4 Å to 2.3 Å. The four compounds (3w32_D1 in blue, Apo in red, 3w33_08 in purple, 3w32_09 in green, and 3w32_10 in orange) had an average RMSD value between 1.5 Å to 2.6 Å. This means that the compound’s value fluctuation falls within the target range, as stated in references (Timofeev et al., 2023; Akash et al., 2023).

We analyzed the tested drugs and controls with respect to their RMSDs to find out which one was more stable. See Figure 8 for the RMSD values for Ligand No. 08 (purple), Ligand No. 09 (green), Ligand No. 10 (orange), and conventional D1 (blue). The compounds’ RMSD was determined after aligning the docking complex with the Apo standard protein backbone. To find the optimal RMSD for each chemical, complex observation was utilized in this situation. The green RMSD of Ligand No. 09 fluctuates, although it stays within the permissible range of 0.5–1.4 Å (Liu et al., 2017; Sherman et al., 2006). Because of minute differences, Ligands No. 08 (purple) and No. 10 (orange) are rather inflexible. There isn't a whole lot of movement in the blue control D1 here. After looking at the RMSD values, this study revealed that all of the ligands were stable, but that Ligand No. 09 (green) was the most stable.

The root mean square fluctuation, or RMSF, can be used to identify and define the local modifications that occur along the protein chain when drugs interact with certain residues. So, to see how the shape of proteins changes when certain ligands attach to a specific remaining site, we calculated and displayed the RMSF value of the initial protein-ligand complexes in Figure 9. The variations of different chemical complexes with the targeted protein 3w32 are compared in this study. Figure 9A compares the 3w32_08 (purple) complex protein to the Apo (red), revealing that only PRO753, HIS870, ARG889, and SER921 were more fluctuating. That is, 98.73% of the residues were stable. Figure 9B compares 3w32_09 (green) complex protein to Apo (red) protein and reveals that THR751, GLU868, GLY873, and SER921 residues are more volatile, implying that 98.73% of residues are stable. Figure 9C compares 3w32_10 (orange) complex proteins to Apo (red), revealing that ALA702, PRO753, and GLY874 residues were more variable, indicating that 99.05% of the residues were stable. The conventional 3w32_D1 complex was compared to Apo (red) protein in Figure 9D, and it was discovered that GLN701, ALA722, GLY873, SER921, and GLY983 residues were more fluctuating, implying that 98.41% of the residues were stable. Protein stiffness is indicated by the substantially smaller variability of residues in the complex structure compared to the native structural components (Apo). Because the N- and C-terminal domains are located at the beginning and end of the protein, the majority of the changes are found there. As a result, ligands 08, 09, and 10 are suitable candidates for a molecule in which the possibility of a specific atom being displaced in a real-world environment is low.

The interactions between the four selected ligands (Ligand 08, Ligand 09, Ligand 10, and Standard D1) and the breast cancer proteins 3w32 were monitored throughout the SID. Drug selectivity, metabolization, and adsorption seem to be significantly influenced by hydrogen-bonding properties in drug design, as demonstrated by the MD simulation’s identification of hydrogen bonds, hydrophobic, ionic, and water bridge interactions. The simulation amply demonstrated the hydrogen bonding connection stated for both compounds up to the very last AA residue. In all structures, the protein residue and the ligand form a variety of interactions, including as hydrogen bonds, hydrophobic interactions, ionic interactions, and water bridges, as Figure 10 shows. The complex 3w32_08 for ligand 08 produced multiple (more than two) interactions at the residues of ASP855, MET793, GLN790, ALA743, ASP800, ASG841, CYS797, LEU844, LEU718, LYS745, PRO794, and ASN842 with an interaction fraction (IF) value of 1.85, 1.25, 1.00, 0.85, 0.55, 0.50, 0.48, 0.45, 0.25, 0.20, and 0.20, respectively. This means that for 155%, 125%, 100%, 85%, 55%, 50%, 48%, 48%, 45%, 20%, and 20% of the simulation time, the specific interaction is maintained by the multiple contacts of the same subtype with the ligand as indicated in Figure 10A. Multiple interactions of the 3w32_09 complex have been observed in the case of ligand 09 at the positions of ASP800 (1.6), MET793 (1.4), ARG841 (1.0), GLN791 (0.98), CYS797 (0.98), THR790 (0.7), SER720 (0.6), ALA743 (0.5), LYS745 (0.45), LEU844 (0.4), LEU718 (0.19), VAL726 (0.13), THR854 (0.10), and ASP855 (0.10) residues maintained by 160%, 140%, 100%, 98%, 98%, 70%, 60%, 50%, 45%, 40%, 19%, 13%, 10%, and 10% of the simulation time the in particular interaction indicated in Figure 10B. For ligand 10, it has been found that multiple interactions of the 3w32_10 complex are maintained by 200%, 198%, 100%, 60%, 50%, 50%, 45%, 30%, 30%, 25%, 15%, and 10% of the simulation time in the specific interaction shown in Figure 10C. These interactions are at the positions of ASP800 (2.00), ASP855 (1.98), MET793 (1.00), ARG841 (0.60), GLN791 (0.50), ASN842 (0.50), LEU844 (0.45), LEU718 (0.30), LYS745 (0.30), THR790 (0.25), THR854 (0.25), and SER720 (0.10) residues. Additionally, it has been discovered that, in the case of the standard D1, multiple interactions of the 3w32_D1 complex are maintained by 285%, 250%, 230%, 210%, 100%, 70%, 100%, and 5% of the simulation time in the specific interaction shown in Figure 10D. These interactions occur at the positions of ASP855 (2.85), PHE856 (2.50), THR854 (2.3), LYS745 (2.1), THR790 (1.00), GLN791 (1.00), ARG776 (0.70), LEU858 (0.10), and CYS775 (0.05) residues.

We assessed the stability of the four compounds (ligand 08, ligand 09, ligand 10, and standard D1) in the MD simulation using ligand characteristics. We employed the Radius of Gyration (rGyr), Intramolecular Hydrogen Bonds (intraHB), Molecular Surface Area (MolSA), Solvent Accessible Surface Area (SASA), and Polar Surface Area (PSA) to examine the characteristics of the ligands. In this analysis the molar surface area (MolSA) as depicted in SF-5 and the Polar Surface Area (PSA) as shown in SF-6, were used to analyze the ligand properties, all of which were found to be favorable for the ligands.

One approach to think about the radius of gyration (Rg) of a protein-ligand complex system is in terms of atomic distribution along its axis. Rg is a useful tool and an important measure of the structural function of a macromolecule that can be used to predict changes in complex stiffness. Thus, we also investigated the stability of four ligands in their interaction with the target protein via Rg throughout the 100 ns simulation time depicted in Figure 11: ligand No. 08 (purple color), ligand No. 09 (green color), ligand No. 10 (orange color), and standard D1 (blue color). The compounds Ligand No. 08, Ligand No. 09, Ligand No. 10, and standard D1 were found to have average Rg values of 3.75, 3.76, 3.25, and 2.4, respectively. This implies that, upon binding the chosen compounds, the protein’s active site did not experience any appreciable conformational changes (Mendichi et al., 2003).

Biological macromolecules’ Solvent-Accessible Surface Area (SASA) affects both their structure and functionality. The residues of amino acids on the surface of proteins are usually hydrophobic or hydrophilic molecules that interact with other molecules and ligands to generate active sites and/or provide information about the behavior of molecules and protein-ligand complexes in various solvents. Thus, Figure 12 displays the SASA value of the protein upon interaction with Ligand Nos. 08, 09, 10, and D1. The complex system’s SASA value, which was found to be a mean between 05 and 165 Å², demonstrated a high level of interaction between an amino acid residue and the selected molecule (Bogatyreva and Ivankov, 2008).

One of the most crucial and significant discoveries made during the SID monitoring for the 100 ns MD simulation is the evaluation of the ligand and protein contract. Following simulations, the four selected ligands (Ligand 08, Ligand 09, Ligand 10, and Standard D1) and the 3w32 proteins identified in breast cancer are shown in their interaction diagrams in Figure 13. At the active sides of ASP855, MET793, GLN791, ASP800, and CYS797, the ligand 08 interacts with protein 3w32 to form numerous (more than two) interactions with simulation times of 99%, 99%, 98%, 45%, and 34% of specific interactions sustained by the multiple contacts illustrated in Figure 13A. When ligand 09 and protein 3w32 are coupled, as Figure 13B illustrates, they interact with the active sides of MET793, GLN791, THR790, ASP800, CYS797, and ARG841 with 88%, 67%, 58%, 42%, 39%, and 35%, respectively. The active sides of ASP800, ASP855, MET793, ASN842, and ARG841 interact at 99%, 99%, 92%, 40%, and 35% when ligand 10 contracts with protein 3w32 in Figure 13C. The active sides LYS745, THR854, ASP855, PHE856, GLN791, ARG776, and THR790 interact at 99%, 99%, 99%, 94%, 57%, 67%, and 49% when the standard D1 contracts with protein 3w32, as shown in Figure 13D.

The development of new medications relies heavily on DMPK analysis. The acronym ADME stands for “absorption, distribution, metabolism, and elimination,” which describes the steps taken by the body to digest and eliminate a medicine (Pellegatti, 2012). Research like this is useful for gauging the possible effectiveness of a drug (Li, 2001). The first is distribution, which describes the speed and extent of medication delivery to various parts of the body following administration. The second component is absorption, which defines the rate and amount of medication absorption into the bloodstream following administration (Yang et al., 2017). The term “elimination” describes how quickly and efficiently a medicine is absorbed into the bloodstream. Metabolic rate (Zhang and Tang, 2018), action mechanism (Currie, 2018), metabolite structure (Guengerich, 2011), and therapeutic efficacy or safety (Smith, 2011) are all aspects of a medicine’s metabolism. The negative and side effects of drugs are commonly referred to as “toxicology”. Table 6 displays the methodology used to acquire data on the medicine’s ADME profile from a computational forecasting internet resource. All possible therapeutic choices are quickly absorbed by humans due to their intestinal absorption rate of about 85%. For therapeutic chemicals to cross the blood-brain barrier, all of them have been found to be subcellularly localized in mitochondria. With values ranging from −1.381 to −6.773, all of the ligands were quite soluble in water; however, ligand 08 had the lowest aqua solubility value and ligand 09 had the highest. Neither a substrate nor an inhibitor for CYP 2D6 could be located.

The possible negative effects of a substance that resembles a drug on living things are referred to as “toxicity” (Smith, 2011). The toxicity of the chosen ligands in Table 7 has been assessed in both aquatic and non-aquatic environments. Furthermore, none of the chemicals have any negative effects on the environment and are safe for ingestion by humans. The industrial compounds’ acute oral toxicity varied greatly, from 1.156 kg/mol to 2.541 kg/mol. With the exception of ligands 05 and 06, all examined ligands have been confirmed to be carcinogen-free, indicating that the chemical described here does not pose a danger of cancer to living things. But it has also been shown that the ligands mentioned are non-toxic, which means that there is no risk to the environment or human health from them. Due to their lack of skin effects, the necessary ligands can be handled freely in the pharmaceutical business.