Preparation and reliability studies of protein structuresThe protein preparation process for structure PDBID: 3H15 resulted in a meticulously refined model, ensuring structural integrity and accuracy. Comprehensive checks for common structural mistakes were conducted, with no errors detected, providing a solid foundation for further optimisation. Metal ions were effectively pre-treated to guarantee proper coordination and stability within the structure. Hydrogens were strategically removed and re-added to optimise hydrogen bonding patterns, enhancing the overall stability of the protein model33,35. The protein preparation documents the meticulous process of preparing the protein PDBID: 5VBN structure. Initially, common structural mistakes were checked for, but none were found. The next step involved pre-treating metals, which was completed. Following this, bond orders were assigned, and no changes were needed. Hydrogens were then removed and re-added, and metals were treated again. Di-sulphur bonds were created, and antibody regions were annotated, with a warning about sequence similarity scores being below a specific cutoff for reference sequences. Selenomethionines were converted, and missing loops were filled using Prime33,35. The structures were adjusted for pH 7.4 and a minimum probability of 0.01, with details provided for each input structure. The process concluded with Epik completing its task, filtering undesired states and idealising hydrogen temperature factors33,37. Protonation penalties were recalculated, and the combined total energy was reported. Various alternative states for specific residues and their respective energy gaps were identified. We optimised hydrogen positions, restrained minimisation, and subsequent refinement levels. The final energy reports were documented, providing insights into bond stretch, angle bending, torsion angle, Lennard Jones, and electrostatic energies. The results show a detailed process of preparing a protein structure with the identifier PDBID: 6NT9. The initial steps involve fixing common structure mistakes, treating metals within the structure, assigning bond orders, and adding or removing hydrogens as necessary. Subsequently, the process moves on to tasks such as creating disulphide bonds, annotating antibody regions, converting Selenomethionines, and filling missing loops using Prime33,35. For each missing loop, the system generates models to fill in the gaps. In this case, there were 10 missing loops, each with multiple models generated. After generating and combining these models, side chains are optimised, and a minimisation process is carried out. The minimisation process involves idealising hydrogen temperature factors and running restrained minimisation using the S-OPLS force field40. Following the refinement steps, the system reports the final energy of the system, potential energy, kinetic energy, temperature, and various energy contributions from the bond stretch, angle bending, torsion angle, 1,4 Lennard Jones, 1,4 electrostatic, Lennard Jones, and electrostatic interactions. The refinement process is iterated multiple times, reporting progress and energy details until reaching a desired refinement level. Various software tools and libraries, such as Schrodinger’s Impact and MMSHARE, perform atom typing, parameter assignments, energy and force calculations, and refinement operations. The results provide a comprehensive overview of the computational steps in preparing and refining the protein structure to achieve an energetically stable conformation suitable for further analysis and simulations. Table 1 comprehensively analyses the energy parameters for three distinct protein structures with PDBIDs: 3H15, 5VBN, and 6NT9. Each structure is assessed based on various energy components crucial for understanding its stability and behaviour. The total energy, comprising both potential and kinetic components, is reported in units of kilocalories per mole (kcal/mol). All structures’ structures, the total kinetic energy and temperature o, indicating a static state typical of molecular modelling simulations. The potential energy, which accounts for the interactions among atoms within the structure, mirrors the total energy values. Specific energy terms shed light on molecular interactions within the protein structures. Bond stretch energy, measured in kcal/mol, represents the energy associated with stretching chemical bonds, while angle bending energy (kcal/mol) reflects the energy required to bend these bonds. Torsion angle energy (kcal/mol) accounts for the energy in twisting the bonds around their axes. Non-bonded interactions are captured by 1,4 Lennard Jones energy (kcal/mol) and 1,4 electrostatic energies (kcal/mol), which describe interactions between atoms within a specified distance cutoff. Additionally, Lennard Jones energy (kcal/mol) and electrostatic energy (kcal/mol) encompass the total non-bonded interactions with short-ures, with short- and long-range contributions (Table 1). Notably, hydrogen bond energy is absent in all cases, indicating either a lack of hydrogen bonds or their exclusion from the energy calculations. This comprehensive energy analysis is treasured into the structural stability and interactions of proteins, providing a foundation for further investigations for drug design. Figure 2 is of 3D representation of prepared protein and the Ramachandran plot for its quality assessment.Table 1 Showing the different energy levels (Kcal/mol) during protein preparations.Fig. 2Showing the 3D structure of the prepared protein with PDBID (A) 3H15, (C) 5VBN and (E) 6NT9, and Ramachandran plot for PDBID (B) 3H15, (D) 5VBN and (F) 6NT9.Protein-ligand interaction analysisProtein-ligand docking or interaction studies in drug design predict how small molecules bind to target protein using computational algorithms and scoring functions by evaluating binding poses and interactions. The interaction between the Droxidopa and Protein MCM10 (PDBID: 3H15) has produced a docking score of − 5.559 Kcal/mol and MM\GBSA score of − 26.04 Kcal/mol and formed two hydrogen bonds in contact with LYS351, LYS353 residues with different OH ligand atom and two salt bridges contact with GLU358, and ARG310 residues with N+H3 ligand’s atom, and O− atom (Fig. 3A,B). It has also generated the potential energy (S-OPLS) is calculated at − 222.474 kcal/mol, with bend energy (S-OPLS) of 424.47 kcal/mol, an LJ-14 energy (S-OPLS) of 852.365 kcal/mol, and dihedral energy (S-OPLS) of 405.957 kcal/mol (Table 2). This indicates a favourable interaction between Droxidopa and MCM10. The negative values for energy parameters suggest stable binding interactions. The high potential and dihedral energy indicate strong molecular interactions, likely contributing to the observed docking and MM/GBSA scores. The Human DNA polymerase epsilon B-subunit (PDBID: 5VBN) has shown a docking score of − 6.835 Kcal/mol and MM/GBSA score of − 37.33 Kcal/mol and formed six hydrogen bonds interactions among ASN491 and N+H3, PHE242, THR245, and SER447 residues with three OH atoms, ASN491 residue with O atom, and formed a salt bridge contact among LYS443 residue and O− atom of the Droxidopa ligand (Fig. 3C,D). The potential energy (S-OPLS) is much higher at − 2637.703 kcal/mol, reflecting a more energetically favourable interaction. The bending energy (S-OPLS) is 2631.907 kcal/mol, LJ-14 energy (S-OPLS) is 5651.879 kcal/mol, and dihedral energy (S-OPLS) is 2915.993 kcal/mol. The docking score is − 6.835 kcal/mol, with an MM/GBSA score of − 37.33 kcal/mol, indicating a strong binding affinity between Droxidopa and the B-subunit of Human DNA polymerase epsilon. The high potential energy and LJ-14 energy suggest significant van der Waals interactions, contributing to the stability of the complex (Table 2). The interaction of TANK-binding kinase 1 (TBK1) (PDBID: 6NT9) showed a docking score of − 6.436 Kcal/mol and an MM\GBSA score of − 20.05 Kcal/mol and has formed seven hydrogen bonds interaction among VAL265 and CYS267 residues with the N+H3 atom, CYS267, LEU269, and ARG427 residues with two OH atoms and TYR427 and ARG427 residues with the O atom of the Droxidopa ligand (Fig. 3E,F). The complex has generated potential energy (S-OPLS) is also high at − 4824.668 kcal/mol, with bend energy (S-OPLS) of 2791.588 kcal/mol, LJ-14 energy (S-OPLS) of 6229.301 kcal/mol, and dihedral energy (S-OPLS) of 2330.94 kcal/mol, indicating a favourable interaction (Table 2)40. The high LJ-14 energy suggests strong van der Waals interactions, contributing to the stability of the complex. The energy parameters provide valuable insights into the stability and strength of the molecular interactions observed in the docking studies. The negative values indicate favourable interactions, with higher negative values suggesting stronger binding affinities. Understanding these energy parameters can help predict the effectiveness of potential drug candidates and optimise their molecular structures for enhanced binding interactions.Table 2 Grid sizes, docking scores, and other computed scores during the molecular docking studies.Fig. 3Showing the 3D ligand-interaction of droxidopa with PDBID (A) 3H15, (C) 5VBN and (E) 6NT9, and 2D ligand interactions of droxidopa with (B) 3H15, (D) 5VBN and (F) 6NT9.Molecular interaction fingerprintsIFPs aid in drug design by capturing and quantifying the unique patterns of molecular interactions between drugs and target proteins. These fingerprints provide insights into binding modes, key interactions, and structural motifs, guiding the design of novel compounds with optimised pharmacological properties and enhanced therapeutic efficacy. The residues that interact with the ligand Droxidopa are crucial in determining its binding affinity and specificity, and those residues are 2VAL, 2LYS, 1ALA, 1ARG, 1ASN, 1CYS, 1GLN, 1GLU, 1ILE, 1MET, 1PHE, 1PRO, 1SER, 1THR (Fig. 4). Ala (Alanine) residues provide hydrophobic interactions, contributing to the drug’s stability within the binding site. Arg (Arginine) residues can form hydrogen bonds with the ligand, enhancing binding affinity through electrostatic interactions. ASN (Asparagine) residues can participate in hydrogen bonding, aiding in specific interactions with the ligand. CYS (Cysteine) residues might form covalent bonds with the ligand, potentially leading to irreversible binding and modulation of drug activity. Gln (Glutamine) residues can engage in hydrogen bonding and hydrophobic interactions, contributing to the stability of the drug-receptor complex. Glu (Glutamic Acid) residues may participate in ionic interactions with the ligand, influencing binding affinity and specificity. Ile (Isoleucine) residues contribute to hydrophobic interactions, stabilising the drug within the binding pocket. Lys (Lysine) residues can form hydrogen bonds and salt bridges, facilitating electrostatic solid interactions with the ligand. Met (Methionine) residues provide hydrophobic interactions and can participate in hydrogen bonding with the ligand. Phe (Phenylalanine) residues contribute to hydrophobic interactions, enhancing the stability of the drug-binding site complex. Pro (Proline) residues may induce conformational changes in the binding site, affecting the orientation and stability of the ligand. Ser (Serine) residues can form hydrogen bonds with the ligand, contributing to specific interactions within the binding pocket. THR (Threonine) residues provide hydrogen bonding and hydrophobic interactions, influencing the binding mode and affinity of the drug. VAL (Valine) residues contribute to hydrophobic interactions, stabilising the drug within the binding pocket (Fig. 4). The diverse interactions between Droxidopa and these residues contribute to its binding affinity, specificity, and pharmacological effects. Understanding these interactions is crucial for optimising the design of drugs targeting the same or similar binding sites, potentially leading to improved therapeutics.Fig. 4Molecular interaction fingerprints of all three protein interacted with the ligand Droxidopa, where the Blue to red colour shows the N to C terminal of the proteins, the upper bars show the interacting residues and the right bars show the count of ligand interactions.DFT and pharmacokineticsDFT computations aid drug design by predicting molecular properties crucial for ligand stability, and analysing HOMO and LUMO levels provides insights into electronic structure and reactivity, guiding the selection of stable ligands. Electron density maps reveal charge distribution, while electrostatic potential maps highlight regions prone to interactions, aiding in ligand optimisation. This comprehensive approach ensures the selection of stable ligands for effective drug design. The DFT computations performed using the Jaguar program reveal significant insights into the properties of Droxidopa as a potential multitargeted inhibitor for cervical cancer proteins. The spin multiplicity of 1 indicates a stable electronic configuration. Employing the UDFT(b3lyp-d3)/SOLV method with a 6–31 g** basis set, the gas phase energy is calculated as − 780.492658 kcal/mol, while the solution phase energy is slightly lower at − 780.52904 kcal/mol. The solvation energy, − 22.83 kcal/mol, highlights the impact of the surrounding environment on the molecule’s stability (Fig. 5A). Relative energy refers to the difference in energy between different conformations or states within a system, providing insights into stability and reactivity. Grad Max and Grad RMS represent the energy gradient’s maximum and root mean square values, respectively, indicating the molecular structure’s magnitude and overall stability, as shown by the proper graphs in Fig. 5A. Disp Max and Disp RMS denote the maximum and root mean square values of the displacement of atoms, offering information on molecular flexibility and conformational changes. Unsigned dE represents the unsigned difference in energy between consecutive optimisation steps, highlighting the convergence of computational methods and the accuracy of results. These parameters collectively impact the reliability and accuracy of computational simulations, guiding the interpretation of molecular properties and behaviour in various environments or biological contexts. Analysis of molecular orbitals shows the alpha HOMO and LUMO at − 0.213568 and − 0.011152, respectively, indicating favourable electronic properties for reactivity (Fig. 5A,B). Similarly, the beta HOMO and LUMO exhibit similar energy levels. Electrostatic potential maps reveal a minimum of − 52.62 kcal/mol and a maximum of 67.88 kcal/mol, indicating regions of attraction and repulsion, respectively. The overall mean electrostatic potential is 3.02 kcal/mol, with a balanced distribution between positive and negative regions. The ALIE analysis further elucidates the molecule’s interaction with its surroundings, with a mean energy of 275.62 kcal/mol. These comprehensive computational results provide valuable insights into Droxidopa’s suitability as a multitargeted inhibitor for cervical cancer proteins, aiding in its further optimisation and development as a potential therapeutic agent (Fig. 5A,B).Fig. 5(A). The DFT computations using the Jaguar Program showed various energy, including the relative energy, which converged around 28 steps. (B) Showing DFT computations using the Jaguar Program showing the HOMO-LUMO (α & β) sites, electron density map and electrostatic potential throughout the Droxidopa structure.Droxidopa, a pharmaceutical compound, exhibits various pharmacokinetic properties determined through computational analysis using tools such as QikProp33,41. The descriptors reveal various characteristics crucial for understanding its behaviour within biological systems. Firstly, considering its molecular structure, Droxidopa comprises 15 atoms, including 6 ring atoms, and forms 15 bonds with 7 rotors. Such structural information provides insights into its overall size and complexity, which can impact its pharmacokinetics. The compound demonstrates solubility with a logarithm of the partition coefficient between octanol and water (QPlogPo/w) of − 2.773, indicating its potential for distribution between lipid and aqueous phases. Additionally, its polar surface area (PSA) of 129.458 square angstroms further characterises its interaction potential with biological membranes and transport proteins. Droxidopa exhibits moderate lipophilicity, as indicated by its logarithm of the octanol-water partition coefficient (QPlogPw) of 14.137, suggesting a tendency for distribution into lipid-rich environments. This is complemented by its QPlogBB value of − 1.63, indicating a slight ability to cross the blood-brain barrier, essential for drugs targeting the central nervous system (Table 3). Furthermore, the compound displays favourable characteristics related to gastrointestinal absorption, with a human oral absorption percentage of 3.026%. This parameter is crucial for understanding orally administered drugs’ bioavailability and potential therapeutic efficacy. In terms of its electronic properties, Droxidopa exhibits molecular polarizability (QPpolrz) of 16.613 cubic angstroms, reflecting its ability to induce temporary dipoles in response to an external electric field. Such polarizability contributes to the compound’s intermolecular interactions and solvation behaviour. Moreover, the compound displays a dipole moment of 2.491 Debye, indicating a degree of charge separation within the molecule, which can influence its interaction with biological targets and transport proteins. In addition to physicochemical properties, Droxidopa also demonstrates characteristics related to molecular geometry and connectivity. For instance, it exhibits an average eccentricity of 6.333 and an average vertex distance degree of 49.333, highlighting the distribution of atoms within the molecule and their spatial relationships (Table 3). The compound’s topological parameters, such as the Balaban distance connectivity index and the connectivity chi-1, provide further insights into its molecular structure and connectivity patterns, which are relevant for understanding its biological activity and pharmacological effects. The comprehensive analysis of Droxidopa’s pharmacokinetic properties reveals its potential as a pharmaceutical agent, with characteristics conducive to absorption, distribution, metabolism, and excretion within biological systems. These insights are valuable for guiding further experimental studies and optimising the compound’s therapeutic utility in clinical applications.Table 3 Showing the pharmacokinetics properties of droxidopa.Molecular dynamics simulationDesmond package in Schrodinger Maestro was used to conduct a 100 nanosecond MD simulation of Droxidopa in a complex with all three protein33,48. Analysis revealed dynamic interactions between Droxidopa and proteins, highlighting potential binding modes and stability over time. This simulation helps understand the behaviour of Droxidopa in complex biological environments. The droxidopa in complex with PDBID: 3H15 has generated 28,910 atoms, the droxidopa in complex with PDBID: 5VBN has generated 73,117 atoms, and the droxidopa in complex with PDBID: 6NT9 has generated 110,088 atoms which were loaded and kept for the production run. The detailed results are as follows-.Root mean square deviationThe Root Mean Square Deviation (RMSD) measures the average distance between the atoms of two superimposed molecules. In the context of molecular dynamics simulations, RMSD quantifies the deviation of atomic positions from a reference structure over time. It is a crucial metric for assessing the stability and convergence of protein-ligand complexes during simulations. Droxidopa in complex with MCM10 protein (PDBID: 3H15) initially deviated to 1.69 Å in the case of protein, while the ligand deviated to 2.30 Å at 0.10 ns. After that, the complex showed a stable performance during the complete simulation period and at 100 ns, protein deviation was 4.01 Å while ligand deviation was noted to be 22.33 Å, and after neglecting the initial deviations, RMSD of protein showed acceptable deviations (Fig. 6A). The human DNA polymerase epsilon B-subunit (PDBID: 5VBN) complex with Droxidopa at the beginning deviated to 1.44 Å for protein, while the ligand to 0.66 Å at 0.10 ns and it keeps deviating and, at 100 ns, the protein deviation was 2.73 Å, and the ligand deviation was 5.44 Å, showing stability after ignoring the initial deviation (Fig. 6B). The TANK-binding kinase 1 (TBK1) (PDBID: 6NT9) in complex with droxidopa initially deviated to 1.60 Å and the ligand to 1.32 Å, at 0.10 ns, and it keeps showing the deviation, and at 100 ns, the protein deviated to 4.19 Å while the ligand deviated to 2.38 Å, and after ignoring the first 1ns complex, it displayed acceptable performance (Fig. 6C). In the case of Droxidopa complexes with MCM10 (PDBID: 3H15), human DNA polymerase epsilon B-subunit (PDBID: 5VBN), and TANK-binding kinase 1 (TBK1) (PDBID: 6NT9), initial deviations were observed. However, after disregarding these, all complexes displayed acceptable stability, with fluctuations in RMSD over time indicating dynamic interactions between the ligand and proteins. Overall, RMSD analysis offers insights into the behaviour and convergence of protein-ligand complexes during simulations.Root mean square fluctuationsRoot Mean Square Fluctuation (RMSF) calculates the average deviation of atomic positions from their average positions throughout a molecular dynamics simulation. It provides insight into the flexibility and dynamics of individual residues within a protein or protein-ligand complex, aiding in identifying regions undergoing significant conformational changes. The complex of MCM10 protein (PDBID: 3H15) with Droxidopa has shown some fluctuating residues beyond 2Å are- GLN235, TYR236, SER252-GLU255, ARG258, LYS259, PRO297-LYS304, LYS353-GLU358, ALA383-GLN393, ASP399, and TYR405-VAL407 and the most interacting residues were ACE234-LYS240, ARG245, LYS248, SER254-ARG258, ARG267, GLN270, LYS293, SER299, ASN301, ASN302, LYS304, PHE306, ARG310, LEU314-LYS319, SER322, PHE324, PHE326, ASP328, LYS331, GLN338, ASN348, MET350-GLU358, SER362, ASP364, VAL376, ASP377, LEU378, THR380-GLN393, and LEU397-VAL407 (Fig. 6D). The complex of Human DNA polymerase epsilon B-subunit (PDBID: 5VBN) with Droxidopa has shown some fluctuating residues beyond 2Å are- HIS84, ASN113, GLU145, PHE147-SER161, THR175-LYS177, GLU192, GLY193, GLU281, SER432-ASN434, GLY526, PHE527, SER2150-PHE2163, LYS2171, SER2174-VAL2181, CYS2187-ALA2192, LYS2223, CYS2224, SER2237-ALA2239, PRO2282 and many of them has interacted with droxidopa are PHE147, HIS214, SER215, TYR218, PHE240, PHE242, PRO243, PRO244, THR245, ASN434, ASN439, LYS443, THR444, SER447, THR490, ASN491, and THR492 (Fig. 6E). The TANK-binding kinase 1 (TBK1) (PDBID: 6NT9) in complex with Droxidopa has shown many fluctuating residues beyond 2Å are- MET1-ASN6, SER12, LYS29-ASP33, ILE43-ARG54, GLU75-HIS81, GLY146-GLN150, GLU165, ASP166, ASP167, SER172, LEU173, MET184-LYS197, PHE224-ARG228, LYS251-GLY255, ASP288, -LYS291, SER328-HIS369, HIS403-ASP409, CYS471-GLN581, and PHE638-LEU658, and the interacting residues are TYR105, GLU109, ASP262-GLY272, GLN274, ALA321, HIS322, LYS323, GLN342, GLY391, LEU392, ILE393, TYR424, ARG427, ILE428, THR431, TYR435 (Fig. 6F). The fluctuations in residues during molecular dynamics simulations can be attributed to several factors, including conformational changes, solvent exposure, and interactions with the ligand or neighbouring residues. Conformational changes in the protein structure, such as loop movements or side-chain rotations, can lead to fluctuations in residue positions. Solvent exposure may also influence residue fluctuations, as residues on the protein surface are more susceptible to solvent interactions and fluctuations. Interactions with the ligand or neighbouring residues can significantly impact residue fluctuations. Residues directly involved in binding interactions with the ligand may experience fluctuations as they adjust their conformations to optimise binding. Additionally, neighbouring residues that indirectly interact with the ligand or undergo allosteric changes due to ligand binding can also exhibit fluctuations. In the case of the MCM10 protein complex with Droxidopa (PDBID: 3H15), fluctuating residues beyond 2Å include regions involved in ligand binding, suggesting dynamic interactions between the protein and ligand. Similarly, in the complexes of human DNA polymerase epsilon B-subunit (PDBID: 5VBN) and TANK-binding kinase 1 (TBK1) (PDBID: 6NT9) with Droxidopa, fluctuating residues are observed, indicating dynamic behaviour of these complexes during simulations. The most fluctuating residues are often located in flexible protein regions, such as loops or termini, where structural changes are more likely to occur. Additionally, residues involved in ligand binding sites or regions undergoing conformational changes due to ligand binding tend to exhibit higher fluctuations.Fig. 6Showing the root mean square deviation (RMSD) of droxidopa in complex with (A) PDBID: 3H16, (B) PDBID: 5VBN, (C) and PDBID: 6NT9 where red shows the ligand deviations, blue shows the Cα deviations and side chains are also shown. Also, the root mean square fluctuations (RMSF) of droxidopa in complex with (D) PDBID: 3H16, (E) PDBID: 5VBN, (F) and PDBID: 6NT9, where blue shows the fluctuation in Cα, and green lines shows the residues interacting the ligand.Simulation interaction diagramsA Simulation Interaction Diagram (SID) visually represents the interactions between a ligand and a protein during molecular dynamics simulations. Various interactions, such as hydrogen bonds, hydrophobic contacts, and electrostatic interactions, are illustrated throughout the simulation. SIDs are valuable for understanding the dynamic behaviour of protein-ligand complexes and identifying key interaction patterns that contribute to binding affinity and stability. The MCM10 protein (PDBID: 3H15) in complex with Droxidopa interacts with many hydrogen bonds among LYS315, ASP316, CYS391, GLN404 residues, and CYS381, ASP318 residues with water molecules along N+H3 atom, THR392, TYR402 residues, and LEU314 residue with water molecule along OH atom and ARG384, CYS391, residues and LYS385 residue with water molecules along two O atom. Also, it forms a salt bridge that contacts ASP316 residue with the N+H3 atom of the ligand (Fig. 7A,B). The Human DNA polymerase epsilon B-subunit (PDBID: 5VBN) in complex with Droxidopa interacts with hydrogen bonds along THR245, PHE242 residues, and SER215, ASN491, THR492 residues with water molecules along three OH atoms and LYS443, ASN491, SER215 residues with two O atoms. Additionally, two pi-cation contacts HIS214 residue with N+H3 atom and LYS443 residue with the benzene ring and form a salt bridge contact LYS443 residue with O atom of the ligand (Fig. 7C,D). Interaction between the TANK-binding kinase 1 (TBK1) (PDBID: 6NT9) and Droxidopa involves fifteen water molecules that act as water bridges. The hydrogen bonds among CYS267, VAL265, LEU269 residue with N+H3 atom, ILE369, ARG271, PRO264, GLN274, GLY391 residues with water molecules interact with three OH atoms and TYR424, ARG427 residues, and SER268, ILE393 residues with water molecules along two O atoms of the Droxidopa ligand (Fig. 7E,F). Analysis of Droxidopa complexes with MCM10, DNA polymerase epsilon B-subunit, and TANK-binding kinase 1 reveals diverse interaction patterns, including the H-bonds, water bridges, pi-cation, and salt bridges contribute to complex stability, highlighting the intricate molecular interactions essential for ligand binding.Fig. 7Showing the simulation interaction diagram (SID) of droxidopa in complex with (A) PDBID: 3H16, (C) PDBID: 5VBN, (E) and PDBID: 6NT9 and the bar graph shows the count of different interaction types of (B) PDBID: 3H16, (D) PDBID: 5VBN, (F) and PDBID: 6NT9, where blue shows water bridges, red shows the ionic interactions, grey shows the hydrophobic and the green shows the H-bonds.MM\GBSA studiesThe Molecular Mechanics Generalised Born Surface Area (MM/GBSA) method is a powerful tool to estimate biomolecular complexes’ binding free energy, such as protein-ligand interactions. By analysing the results obtained from MM/GBSA calculations performed using Schrodinger Maestro, where 100 nanosecond molecular dynamics (MD) simulations were conducted for each protein-ligand complex, we can gain insights into the performance of the method as the number of frames increases. As the number of frames increases from 0 to 1000, the MM/GBSA method provides a comprehensive understanding of the energetics of protein-ligand binding (Fig. 8). At frame 0, the complex energies for all three complexes, 3H15, 5VBN, and 6NT9, are relatively high, indicating unfavourable interactions between the protein and ligand. This is reflected in the negative binding free energy values, suggesting weak or non-existent binding between the molecules. As the simulation progresses, the complex energies fluctuate, reflecting the dynamic nature of the protein-ligand complexes within the simulated environment. Notably, for complex 3H15, there is a gradual decrease in complex energy over the first few frames, indicating a stabilisation of the protein-ligand complex (Fig. 8). This is accompanied by a corresponding increase in the binding free energy, suggesting a strengthening of the protein-ligand interaction. However, for complexes 5VBN and 6NT9, the complex energies remain relatively high throughout the simulation, indicating persistent unfavourable interactions between the protein and ligand. Despite fluctuations in energy values, there is no clear trend towards stabilising or strengthening the protein-ligand complex. Overall, the performance of the MM/GBSA method varies depending on the specific protein-ligand complex analysed. While it demonstrates the ability to capture dynamic changes in the energetics of protein-ligand interactions, its effectiveness in predicting binding affinities may be limited by the complexity of the molecular system and the accuracy of the force field parameters used in the simulations. In conclusion, the MM/GBSA method provides valuable insights into the energetics of protein-ligand binding, but its performance can vary depending on the specific molecular system studied. Further refinement and validation of the method may be necessary to improve its predictive accuracy for drug discovery and design applications.Fig. 8Showing the molecular mechanics of generalised born surface area (MM/GBSA) studies conducted on MD simulation trajectories. The blue bard shows the total complex energy, whereas the orange shows the binding free energy of the complex.
