Skip to main content
NCBI home page
In Silico Pharmacology logoLink to In Silico Pharmacology
. 2026 Jan 14;14(1):36. doi: 10.1007/s40203-025-00542-y

Molecular insights into the bioactivity of H-thiazine compounds against breast cancer cells: a computational study

Lesego M Mogoane 1, Vincent A Obakachi 1, Penny P Govender 1, Krishna K Govender 1,✉
PMCID: PMC12804472  PMID: 41552637

Abstract

Breast cancer, a leading cause of global mortality, necessitates novel therapies targeting key drivers like the epidermal growth factor receptor (EGFR). This computational study evaluates nine 4-phenyl-2H-[1,3]thiazino[3,2-a]benzimidazol-2-imine (H-thiazine) derivatives as potential EGFR inhibitors. Using molecular docking, ADMET profiling, molecular dynamics simulations, and binding energy calculations, we identified methyl- and bromine-substituted derivatives as probable candidates that outperform the reference drug Olmutinib in terms of binding affinity, pharmacokinetics, and stability. Although these compounds showed promising bioactivity, in silico toxicity screening indicated potential AMES mutagenicity and hERG-II inhibition, highlighting important safety liabilities. Overall, thiazine derivatives represent viable scaffolds for EGFR-targeted anti-cancer development; however, further optimization and experimental validation, including biochemical assays and genotoxicity testing, are required to confirm their therapeutic potential.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40203-025-00542-y.

Keywords: Breast cancer, EGFR, Thiazine derivatives, Computational drug design, Molecular dynamics

Introduction

Breast cancer is a critical threat to global health, with 2.3 million new diagnoses and 670,000 deaths reported worldwide in 2022(World Health Organization 2022). In Sub-Saharan Africa, 146,130 cases and 71,662 deaths were recorded, with projections indicating an 85.7% increase in new cases and a 89% rise in deaths by 2040(Sung et al. 2021). Current treatments, including surgery, chemotherapy, radiotherapy, hormone therapy, and targeted therapies, often cause severe side effects such as hair loss, nausea, fatigue, and increased infection risk, as well as long-term complications like heart problems and secondary cancers (National Cancer Institute 2023). Resistance to these therapies further limits their effectiveness, highlighting the need for more targeted and efficient treatment options developed for a specific breast cancer subtype (Waks and Winer 2019).

A promising strategy involves targeting molecular drivers like the epidermal growth factor receptor (EGFR), a member of the Epidermal growth factor receptor (ErbB) family of receptor tyrosine kinases, which is overexpressed in aggressive breast cancers and associated with poor clinical outcomes (Yarden and Sliwkowski 2001; Masuda et al. 2012). EGFR, along with HER2, HER3, and HER4, features an extracellular domain, a transmembrane segment, and an intracellular tyrosine kinase domain that regulates cell growth and survival, as shown in Fig. 1 (Hynes and Lane 2005; Berasain et al. 2009). Inhibitors such as gefitinib, cetuximab, and lapatinib target EGFR but face challenges including resistance and suboptimal pharmacokinetics (Baselga and Arteaga 2005; Tebbutt et al. 2013). Developing next-generation EGFR inhibitors with improved selectivity and pharmacokinetics (Sharma et al. 2007) is crucial for enhancing efficacy and reducing adverse effects.

Fig. 1.

Fig. 1

Open in a new tab

ErbB (HER) family (Ismail et al. 2019)

Computational tools, particularly structure-based drug design (SBDD), facilitate drug discovery by leveraging 3D protein structures to design potent and selective compounds (Anderson 2003; Kalyaanamoorthy & Chen 2011). These tools, as shown in Fig. 2, are critical for identifying novel chemical scaffolds with therapeutic potential, particularly thiazine compounds.

Fig. 2.

Fig. 2

Open in a new tab

In Silico approaches to drug discovery and development

Thiazines, a class of heterocyclic molecules found in natural and synthetic compounds, exhibit diverse biological activities, including antibacterial, anti-cancer, anti-inflammatory, antitumor, and antiviral properties (Tilak 1960; Sanphanya et al. 2013; Sharma and Makkar 2016; Badshah & Naeem 2016). Sulphur and nitrogen-containing heterocycles have been widely studied for their anti-cancer, antimicrobial, and anti-inflammatory properties, attributed to their ability to engage in diverse hydrogen bonding and π–π stacking interactions within enzyme active sites (Kumar et al. 2024; Wang et al. 2021; Hamad 2025). Despite their promise, thiazine derivatives remain underexplored as EGFR inhibitors for breast cancer.

The parent scaffold 4-phenyl-2H-[1,3]thiazino[3,2-a]benzimidazol-2-imine (H-thiazine) offers an attractive foundation for rational optimization, particularly through small substituents such as methyl and halogens that can modulate electronic distribution, lipophilicity, metabolic stability, and binding affinity (Chiodi & Ishihara 2024; Shi et al. 2022). These substituents have been shown to enhance receptor interactions while maintaining drug-like molecular properties. Given the chemical flexibility and pharmacological potential of the H-thiazine scaffold, a systematic computational evaluation is warranted.

In this study, nine thiazine derivatives (Fig. 3) were evaluated, including the parent H-thiazine scaffold, as potential EGFR inhibitors. Using in silico approaches, including molecular docking, ADMET profiling, molecular dynamics simulations, and MM-GBSA binding free energy calculations, this study assesses the binding affinity, pharmacokinetic properties, dynamic stability, and therapeutic potential of these compounds against EGFR in breast cancer. This work presents the first comprehensive computational assessment of thiazine-based derivatives against EGFR, highlighting promising candidates for further optimization and future experimental validation.

Fig. 3.

Fig. 3

Open in a new tab

Chemical structures of H thiazine and its derivatives (A–I). A Compound H, B Compound Br, C Compound M-M1, D Compound Cl, E Compound F, F Compound Nitro, G Compound Methyl, H Compound P-methoxy, and I Compound M-M2. Substituent positions for each derivative are indicated at C17 for compounds H, Br, Methyl, Cl, F, P-methoxy, Nitro, and C16 for compounds M-M1 and M-M2

Methods

Ligand preparation

This study investigated nine thiazine-based derivatives, including 4-phenyl-2H-[1,3]thiazino[3,2-a]benzimidazol-2-imine (H-thiazine), which were optimized at the DFT/M06-2X level with the aug-cc-pVTZ basis set to ensure geometric and electronic stability prior to further analysis (Kendall et al. 1992; Zhao and Truhlar 2008). Olmutinib, an FDA-approved EGFR inhibitor, was included as the reference compound, with its 3D structure retrieved from the PubChem database (PubChem CID: 71461008) (Kim et al. 2023). All ligands were prepared using the LigPrep module in Maestro 13.6 version of the Schrödinger Suite 2023-2, involving hydrogen atom addition and geometry optimization using the OPLS-2005 force field (Banks et al. 2005; Schrödinger 2023a). During ligand preparation (LigPrep), tautomer generation was limited to the most likely tautomer for each compound. While multiple tautomers are chemically possible, the focus was on the dominant physiological form to maintain consistency and reduce computational redundancy. The ligand protonation states were set to pH 7.0 ± 2.0 using Epik 2 (Obakachi et al. 2025). The protonation states were set based on physiological pH, and the predicted pKa values of functional groups were used to justify these assignments. The processed structures were saved for molecular docking calculations.

Protein preparation

The EGFR tyrosine kinase domain crystal structure (PDB ID: 1M17) was retrieved from the Protein Data Bank (RCSB Protein Data Bank 2024) and prepared using Schrödinger's Maestro suite (Baby et al. 2020; Schrödinger 2023a; Shri et al. 2024). The protein preparation involved removing water molecules and heteroatoms, followed by the addition of hydrogen atoms to the protein structure, and optimizing the structure (Bhargavi et al. 2025). After protein preparation, the structure was split into protein, ligand, and water using the Split Protein function in Maestro. This step ensured that all non-protein components, such as water molecules and the co-crystallized ligand, were separated into independent entries within the workspace navigator. This allowed precise selection of components during receptor grid generation. After this step, the prepared protein structure was finalized and ready for grid generation and docking.

Molecular docking

Ligand docking is a computational process that predicts how a ligand (a small molecule) binds to a protein (receptor) and estimates their interactions (Wani et al. 2025). The EGFR binding site was defined based on the co-crystallized ligand (erlotinib), with default box dimensions ensuring full coverage of the active site. A 3D receptor grid was then generated around the active site to define the docking region. The grid box encompassed all key catalytic residues to ensure accurate placement of the ligand within the binding pocket. This grid limits the search space to the relevant area, focusing docking simulations on potential interaction sites (Ottu et al. 2025). The Glide XP docking methodology, known for its high precision, was employed to investigate ligand–protein interactions (Naidoo et al. 2025). The OPLS-2005 force field refined docking by accounting for ligand movements (Gunasekaran and Dhakshinamurthy 2024). For each ligand, up to 5 poses were generated and ranked using Glide XP scoring. The docked conformers were evaluated using the docking score, providing valuable insights for future drug design.

ADME studies

The docked ligands underwent ADME analysis using QikProp in Maestro software to predict pharmacokinetic properties and assess drug-likeness (Schrödinger Release 2023a, b-2 2023a). The analysis was run in fast mode and included identifying the five most similar known drug molecules for comparison (Ali and Al-Hamashi 2024).

Molecular dynamics simulation

Molecular dynamics (MD) simulations were performed on the EGFR-ligand complexes (PDB ID: 1M17) using the Desmond MD simulation package within the Schrödinger 2023a, b-2 suite, with the OPLS-2005 force field (Schrödinger Release 2023a, b-2, 2023b). The simulations were carried out for selected thiazine derivatives and the reference compound Olmutinib to evaluate the dynamic stability of the protein–ligand interactions (Alzain et al. 2025). Each protein–ligand complex was placed in an orthorhombic simulation box with a 10 Å buffer zone between the solute and the box edges, and solvated using the TIP3P water model. Na⁺/Cl⁻ ions were added to neutralize the system and achieve a physiological 0.15 M ionic concentration, resulting in approximately 47,559–50,615 atoms per system. The OPLS-2005 force field was used for the protein, and ligand parameters were derived automatically using the OPLS-2005 force field. All bond lengths were unconstrained. Coulombic interactions were used for electrostatics, and a 9.0 Å cutoff was applied for nonbonded interactions. The systems were energy-minimized using the steepest descent algorithm, followed by equilibration and production in the NPT ensemble at 300 K and 1 atm for 200 ns. A Nose–Hoover thermostat and a Martyna-Tobias-Klein barostat were used with default relaxation times. No positional restraints were applied. Trajectory data were saved at 20 ps intervals, resulting in 2000 frames per simulation for post-simulation analysis (Gheidari et al. 2024).

Binding free energy calculations

The binding free energy (ΔGbind) is a more reliable criterion for ranking ligands based on their true binding affinity compared to empirical docking scores, such as the XP docking score. It reflects a combination of molecular mechanical energy and solvation contributions (both polar and nonpolar), providing a more physically realistic estimate of ligand-receptor interaction strength (Abbasi et al. 2024). In this study, MM-GBSA calculations were conducted using a custom Python script (thermal_mmgbsa.py) executed via the MobaXterm software (Yazdani et al. 2024). The MM-GBSA was performed on snapshots extracted from 200 ns of the MD simulations (Fereidounpour et al. 2024). The trajectory files were first opened and visually inspected in Maestro to ensure the protein–ligand complex was equilibrated and to determine the total number of frames. Snapshots from frames 1 to 2000 were extracted at an interval of 10. The script was submitted via the terminal. After the calculations were completed, the output files were opened and read to obtain the binding energy results (Mbayo et al. 2025). The average ΔGbind was recorded in kcal/mol, and more negative ΔG_bind values indicate stronger binding. This approach allowed automated processing of molecular dynamics trajectories to compute binding free energies for each ligand–EGFR complex.

Toxicological profile

To assess the safety of the lead compounds, an in silico toxicology analysis was conducted using the pkCSM web server (PKCSM). The SMILES notation for each compound was used to predict potential toxicity (Dulsat et al. 2023).

Results and discussion

Validation of docking protocol

The accuracy of the molecular docking methodology was rigorously validated by re-docking the co-crystallized ligand within the EGFR active site and comparing its predicted conformation to the original crystal structure, as illustrated in Fig. 4 with the RMSD calculated based on the heavy-atom alignment between the re-docked pose and the crystallographic ligand. The Root Mean Square Deviation (RMSD) between these conformations was determined to be 1.61 Å, a value comfortably below the widely accepted threshold of 2.00 Å (Mukherjee et al. 2010; Zev et al. 2021; Llop-Peiró et al. 2024). This low RMSD reflects a high degree of congruence between the computational predictions and experimental data, affirming the docking protocol's precision in modeling ligand-receptor interactions. In line with this, Zare et al. (2023) reported low RMSD values for 6-bromoquinazoline derivatives, indicating accurate reproduction of experimental ligand poses. This validation is pivotal, as it bolsters confidence in the reliability of predicted binding poses for novel ligands, ensuring alignment with empirical observations and enhancing the robustness of subsequent docking studies for drug discovery applications (Meng et al. 2011).

Fig. 4.

Fig. 4

Open in a new tab

Superimposed structures of co-crystallized and re-docked reference ligand

Molecular docking and binding affinity analysis

To assess the therapeutic potential of nine thiazine derivatives as EGFR inhibitors, molecular docking and MM-GBSA binding free energy calculations were conducted, with olmutinib, a well-established FDA-approved inhibitor, serving as the reference standard (Elsebaie et al. 2024). All binding free energy (ΔG_bind) values were reported in kcal·mol⁻1 for consistency, using the standard conversion factor (1 kcal/mol = − 4.184 kJ/mol).

The results, detailed in Table 1, reveal that olmutinib achieved a docking score of − 4.93 kcal/mol and a binding free energy (ΔGbind) of -12.13 kcal/mol, providing a baseline for comparison. Among the thiazine derivatives, the methyl-substituted compound emerged as a standout, exhibiting the most favourable docking score of − 6.95 kcal/mol and a ΔGbind of − 12.86 kcal/mol, indicating superior binding affinity and structural complementarity with the EGFR active site. Similarly, the nitro derivative demonstrated exceptional performance, with a docking score of − 6.32 kcal/mol and the most negative ΔGbind of − 13.19 kcal/mol, highlighting its potent binding energetics.

Table 1.

Molecular docking and binding free energy results

Compounds Docking score kcal/mol Binding free energy kcal/mol
Olmutinib − 4.93 − 12,13
H − 6.62 − 11.56
Br − 5.47 − 12.56
Cl − 5.81 − 10.26
F − 5.56 − 10.18
P-methoxy − 4.87 − 11.63
Nitro − 6.31 − 13.19
Methyl − 6.95 − 12.86
Open in a new tab

The bromine (Br) derivative presented a balanced profile, with a docking score of − 5.47 kcal/mol and a ΔGbind of − 12.56 kcal/mol, underpinned by robust hydrophobic interactions that contribute to a stable ligand-receptor complex. In contrast, the hydrogen (H) derivative, despite a promising docking score of − 6.62 kcal/mol, recorded a less favourable ΔGbind of − 11.56 kcal/mol compared to olmutinib. The fluorine (F) and chlorine (Cl) derivatives, with docking scores of − 5.56 kcal/mol and -5.81 kcal/mol, respectively, showed moderate binding potential but weaker ΔGbind values (− 10.18 and − 10.26 kcal/mol), indicating fewer effective interactions with EGFR. Collectively, the methyl, nitro, and bromine derivatives outperformed olmutinib, positioning them as promising leads for further development into next-generation EGFR inhibitors tailored for breast cancer treatment.

Protein–ligand interaction analysis

Molecular docking provides critical insights into the interactions between ligands and the EGFR protein (PDB ID: 1M17), highlighting key binding features such as hydrogen bonds, pi-pi stacking, hydrophobic contacts, electrostatic interactions, and polar contacts that influence binding affinity and stability (Kitchen et al. 2004; Pan et al. 2022). These interactions, visualized across the multi-panel Fig. 5, reveal a consistent engagement of core residues, including LEU768, MET769, GLN767, and LYS721, across all nine thiazine derivatives and the reference compound Olmutinib, underscoring a shared binding scaffold within the EGFR active site.

Fig. 5.

Fig. 5

Fig. 5

Open in a new tab

Protein–Ligand Interaction of A Olmutinib, B Compound H, C Compound Br, D Compound Cl, E Compound P-methoxy, F Compound F, G Compound methyl, and H Compound nitro with amino acid residues of EGFR (1M17). The compounds form hydrogen bonds primarily with MET769, while π-cation interactions are observed with LYS721. Hydrophobic contacts with residues such as LEU768, THR766, and PHE699 further stabilize the complex

Olmutinib exhibits moderate binding, characterized by hydrophobic interactions with LEU768 and PHE699, supplemented by a single hydrogen bond with MET769, reflecting a balanced but less optimized interaction profile. The thiazine derivatives, however, display varying degrees of enhancement. Compound h mirrors olmutinib's pattern, with hydrophobic contacts to PHE699 and LEU768, reinforced by a hydrogen bond with MET769; however, its more compact orientation suggests an improved fit. Compound br enhances this profile with additional hydrophobic interactions involving MET742 and MET769, as well as polar contacts with LYS721 and GLN767, indicating greater stability. In contrast, compound cl relies heavily on hydrophobic engagements with PHE699 and LEU768, with weaker polar interactions at LYS721 and potential dipole contributions from GLN767 and THR766, lacking strong hydrogen bonds. Compound p-methoxy and compound f show similar trends, with stable hydrophobic embedding via GLU738, THR766, and LEU768 (p-methoxy) and MET769, LEU768, and PHE699 (f), respectively, alongside weaker polar interactions with LYS721 and GLN767. These interactions suggest a dipole potential but fall short of robust electrostatic bonding. Compound methyl stands out due to its extensive hydrophobic contacts with MET769, LEU768, and PHE699, complemented by a strong hydrogen bond with MET769, which aligns with its superior docking and MM-GBSA scores. Compound nitro excels further, featuring a rich interaction network that includes a hydrogen bond to MET769, strong ionic interactions between the oxygen atoms of its nitro group and GLU738 and LYS721, as well as hydrophobic contacts with PHE699, CYS773, and LEU768, which reflects its exceptional binding energetics.

The standout performers, methyl, br, and nitro, distinguish themselves through more diverse and robust interaction profiles, including stronger hydrogen bonds and electrostatic contributions, which correlate with their favourable docking scores and binding free energies. This enhanced complementarity with the EGFR active site positions these derivatives as promising leads for further optimization, while the varying interaction strengths across the series highlight opportunities for structural refinement to boost efficacy.

All the thiazine compounds analysed were found to form a pi-cation interaction with LYS721, a positively charged amino acid situated in the ATP-binding region of EGFR. According to Palanivel et al. (2021), this residue plays a crucial role in kinase inhibition, as its interaction with inhibitors helps stabilize ligand binding and prevents ATP from entering the catalytic site. Similarly, Türkmenoğlu (2022) emphasized that pi-cation interaction with LYS721 is a recurring feature among potent EGFR inhibitors, contributing directly to their inhibitory activity.

ADME analysis

The ADME properties of nine thiazine derivatives and the reference compound olmutinib were predicted using the QikProp module, providing a comprehensive assessment of their absorption, distribution, metabolism, and excretion profiles (Schrödinger 2023b). All compounds adhered to the Lipinski Rule of Five (Ro5), demonstrating no violations and confirming their suitability for oral drug-likeness, characterized by molecular weights below 500 g/mol, balanced lipophilicity (Log P < 5), and favourable hydrogen bond donor and acceptor counts that enhance absorption and permeability, as detailed in Table 2. Olmutinib, despite being an approved drug, exhibited limitations due to poor solubility, an elevated solvent-accessible surface area (SASA) of 831 Å2, and reduced permeability, which is consistent with its known pharmacokinetic challenges. In contrast, the thiazine derivatives exhibited more compact structures with SASA values ranging from 500 to 540 Å2, facilitating improved diffusion across membranes.

Table 2.

ADME results

Properties Optimum value H Nitro Methyl F Br P-methoxy Cl Olmutinib
MW  < 500 277 322 291 295 356 307 312 487
CNS − 2 (Inactive) + 2 (Active) 0 0 − 2 0 0 0 0 1
Dipole 1–12.5 5.2 2.1 5.8 2.8 3.4 4.9 3 2.9
SASA 300–1000 503 543 534 512 531 524 521 831
DonorHB 0–6 1 1 1 1 1 1 1 2
AccptHB 2–20 2.5 3.5 2.5 2.5 2.5 3.5 2.5 8.5
QPlogPo/w − 2–6.5 3.6 2.9 3.9 3.9 4.2 3.7 4.1 4.5
QPlogS − 6.5–0.5 − 4.4 − 4.6 − 5.0 − 4.8 − 5.3 − 4.4 − 5.0 − 6.2
QPPCaco  < 25 poor, > 500 great 1556 185 1557 1562 1562 1597 1593 320
QPlogBB − 3–1.2 − 0.195 − 1.19 − 0.220 − 0.088 − 0.027 − 0.230 − 0.015 − 0.358
QPP-MDCK  < 25 poor, > 500 great 1410 141 1411 2558 3757 1450 3560 277
QPlogKhsa − 1.5–1.5 0.39 0.35 0.54 0.43 0.53 0.39 0.50 0.73
Metab 1–8 1 2 2 1 1 2 1 3
%HOA  < 25% poor, > 80% high 100 84.8 100 100 100 100 100 100
PSA 7–200 41.4 86.2 41.3 41.3 41.3 49.8 41.3 84.7
RO3 0 0 0 0 0 0 0 1
RO5 0 0 0 0 0 0 0 0
Open in a new tab

MW (Molecular Weight), CNS (Central Nervous System activity/potential), Dipole (Dipole moment), SASA (Solvent Accessible Surface Area), DonorHB (Hydrogen Bond Donors), AccptHB (Hydrogen Bond Acceptors), PSA (Polar Surface Area). Lipophilicity and solubility: QPlogPo/w (Predicted octanol/water partition coefficient), QPlogS (Predicted aqueous solubility). Permeability and absorption: QPPCaco nm·s⁻1 (Predicted Caco-2 cell permeability), QPlogBB (Predicted blood–brain barrier permeability), QPP-MDCK nm·s⁻1 (Predicted MDCK cell permeability), %HOA (Percent Human Oral Absorption). Metabolism and binding: Metab (Number of metabolic sites), QPlogKhsa (Predicted binding to human serum albumin). Rule-based evaluations: RO3 (Rule of Three), RO5 (Rule of Five) (Kumar and Ayyannan 2023.

Compounds h, methyl, br, f, p-methoxy, and cl demonstrated exceptional membrane permeability, with Caco-2 and MDCK values exceeding 1400 and 100% predicted human oral absorption, markedly outperforming the lower permeability observed in nitro and olmutinib. The predicted human oral absorption (%HOA) of 100% is an estimate provided by QikProp based on calculated permeability and physicochemical properties; it does not represent experimentally measured absorption. Notably, nitro distinguished itself by exhibiting poor blood–brain barrier penetration and reduced permeability, although this is offset by its high binding affinity, suggesting a selective profile. Most thiazine derivatives, excluding nitro and olmutinib, maintained lower polar surface areas (PSA), enhancing permeability potential. Additionally, these derivatives exhibited greater metabolic stability, with predicted metabolic sites limited to 1–2, compared to olmutinib, which has three. All thiazine compounds complied with Jorgensen's Rule of Three (Ro3), whereas olmutinib breached one criterion due to its suboptimal Caco-2 permeability of 320.

Overall, the thiazine derivatives, particularly h, methyl, and cl, showcased superior pharmacokinetic characteristics compared to olmutinib, including enhanced permeability, oral absorption, and metabolic stability. These attributes position them as promising candidates for further optimization as orally bioavailable EGFR inhibitors.

However, it is essential to note that ADMET predictions are inherently dependent on the quality and scope of the experimental data used to train the predictive models and therefore serve as theoretical estimations rather than definitive pharmacokinetic measurements (Kesharwani et al. 2020). Similarly, molecular docking scores may not fully capture real binding energetics due to simplified scoring functions that overlook entropic and solvation effects as well as non-classical interactions such as halogen and guanidine-arginine bonding (Sethi et al. 2020). To mitigate these limitations, molecular dynamics (MD) simulations and MM-GBSA binding energy analyses were performed, confirming the dynamic stability and favourable binding behaviour of the top thiazine–EGFR complexes under physiological conditions.

Molecular dynamics simulation and post-MD analysis

Molecular dynamics (MD) simulations were conducted to evaluate the stability and dynamic behaviour of ligand-EGFR complexes over a 200 ns timeframe, utilizing key metrics such as Root Mean Square Deviation (RMSD), Root Mean Square Fluctuation (RMSF), Solvent-Accessible Surface Area (SASA), and Radius of Gyration (rGyr) (Adcock and McCammon 2006). RMSD measures the overall deviation from the initial structure, while RMSF assesses residue-specific flexibility, offering valuable insights into binding stability. Based on prior docking and binding energy analyses, the top-performing thiazine derivatives methyl, bromo, and nitro were selected for detailed examination in the main manuscript, with their RMSD and RMSF profiles presented in Fig. 6A–H. Data for the remaining derivatives (f, p-methoxy, cl, h) are provided in Supplementary Figs. S1 and S2 to support a comprehensive analysis.

Fig. 6.

Fig. 6

Open in a new tab

Protein–ligand RMSD and RMSF for A, B Methyl thiazine derivative, C, D Bromo thiazine derivative, E, F Nitro thiazine derivative, G, H Olmutinib over 200 ns

The RMSD profiles revealed notable differences in stability among the compounds. From Fig. 6 and Table 3, the hydrogen (h) derivative exhibited the lowest RMSD at 1.18 Å, followed by br at 1.77 Å, fluorine (f) and chlorine (cl) at 2.59 Å and 2.80 Å, respectively, and methyl at 2.91 Å, indicating minimal structural deviation and robust maintenance of their initial binding poses within the EGFR active site. olmutinib, with an RMSD of 4.47 Å, displayed significant instability and fluctuating binding, likely due to a suboptimal fit. Nitro, with an RMSD of 3.47 Å, showed moderate stability, aligning with its strong initial binding energy but suggesting dynamic variability over time. The nitro derivative showed moderate RMSD but a highly favourable MM-GBSA binding energy (-55.19 kJ/mol), indicating that despite some conformational flexibility, strong interactions with key residues were maintained. In contrast, compound h exhibited the lowest RMSD (1.18 Å), reflecting excellent structural stability, but its binding energy was lower than the reference, suggesting weaker thermodynamic interactions. These results illustrate that structural stability alone does not guarantee strong binding, and that combining RMSD and MM-GBSA analyses provides a more complete understanding of ligand performance.

Table 3.

Summary of molecular dynamics simulation results

Compounds RMSD (Å) RMSF (Å) rGyr (Å) SASA(Å2)
F 2.59 2.24 3.57 58.10
P-methoxy 3.95 3.20 3.98 67.10
Br 1.77 1.50 3.66 49.20
Cl 2.80 1.39 3.57 94.10
Nitro 3.47 2.52 3.74 55.20
H 1.18 0.80 3.59 73.30
Methyl 2.91 2.33 3.88 53.00
Olmutinib 4.47 1.78 5.75 285.0
Open in a new tab

The compounds that maintained low RMSD values (< 2 Å) demonstrated strong and stable binding within the EGFR active site, exhibiting minimal structural deviation throughout the simulation. In line with the findings of Agarwal et al. (2022), who observed stable RMSD profiles for natural product inhibitors over a 100-ns trajectory, these results indicate that the ligands remained consistently bound to the protein throughout the entire simulation period.

RMSF analysis further illuminated these trends, with br at 1.50 Å and h at 0.80 Å reflecting reduced residue flexibility in the active site, indicative of tighter binding. In agreement with this, Emami et al. (2022) reported that lower RMSF values are associated with reduced atomic fluctuations and increased structural compactness of the protein, which collectively indicate greater stability of the protein–ligand complex. In contrast, olmutinib's RMSF of 1.78 Å and nitro's 2.52 Å indicated greater local fluctuations, suggesting less conformational stability (Van Der Spoel et al. 2005).

SASA and rGyr (Table 3) analyses provided more profound insights into the compactness of the complexes. Br and methyl recorded the lowest SASA values at 49.20 Å2 and 53.00 Å2, respectively, signifying deep embedding within the EGFR pocket and enhanced stability. At the same time, olmutinib's SASA of 285.0 Å2 suggested a more exposed, less tightly bound conformation. Nitro's SASA of 55.20 Å2 indicated a moderately stable binding mode, although less optimal than that of the top performers. The rGyr values, as shown in Table 3 for br (3.66 Å), h (3.59 Å), and methyl (3.88 Å), were lower and more stable, reflecting compact and conformationally intact complexes. In contrast, olmutinib's rGyr of 5.75 Å and p-methoxy's 3.98 Å correlated with less stable binding dynamics. The alignment of SASA and rGyr data highlights the superior dynamic stability of methyl, br, and nitro groups, with the thiazine derivatives collectively demonstrating greater resilience and compactness compared to Olmutinib, as illustrated in Fig. 7 (Bakan et al. 2011). The thiazine compounds maintained stable structural compactness throughout the simulation, exhibiting only minor fluctuations compared to olmutinib, as shown in Fig. 7. This stability indicates strong binding and minimal conformational changes, suggesting robust and durable interactions with the EGFR protein. The consistently low radius of gyration (rGyr) values (below 4 Å) observed for the thiazine EGFR complexes further confirm a compact and stable conformation. These findings are consistent with previous reports, such as Balkrishna et al. (2024), who demonstrated that lower rGyr values correlate with enhanced protein–ligand stability, as observed in the luteolin HSP90 complex.

Fig. 7.

Fig. 7

Open in a new tab

Radius of gyration dynamics over 200 ns molecular dynamics simulation

These results underscore the potential of methyl, br, and nitro as lead candidates, supported by their stable binding profiles. At the same time, supplementary data for the other derivatives offer a foundation for future refinement.

Protein–ligand contact analysis

The chemical nature of ligand–protein interactions within the EGFR active site was further elucidated, with hydrophobic interactions emerging as the dominant force, complemented by significant contributions from van der Waals forces and π-π stacking (Goodsell and Olson 1990). This analysis, illustrated in Fig. 8 for the top-performing thiazine derivatives methyl, bromo, and nitro, alongside the reference compound olmutinib, highlights their binding characteristics. Hydrophobic ligands, such as methyl and bromine, thrive when deeply embedded within the protein pocket, where low Solvent-Accessible Surface Area (SASA) values, such as 53.0 Å2 for methyl and 49.2 Å2 for bromine, correlate with enhanced binding stability and favorable interactions. This burial within the hydrophobic core highlights the crucial role of these interactions in stabilizing a ligand-EGFR complex.

Fig. 8.

Fig. 8

Open in a new tab

Protein–ligand contact for A Methyl derivative, B Bromo derivatives, C Nitro derivative and D Olmutinib

In contrast, polar ligands like nitro and olmutinib rely on hydrogen bonds and electrostatic interactions to anchor their binding, as depicted in Fig. 8C, D. While nitro benefits from ionic interactions with residues such as GLU738 and LYS721, and olmutinib forms a hydrogen bond with MET769, their effectiveness depends on precise orientation within the binding site, with some tolerance for solvent exposure.

The hydrophobic residues VAL702, ALA719, and LEU820 identified in this study align with those reported as key contributors to ligand binding and stability within the EGFR kinase domain in previous research. Similarly, Moussaoui et al. (2024) highlighted the significance of these residues, showing that quinazoline derivatives maintained stable van der Waals and hydrophobic interactions with VAL702, ALA719, and LEU820 throughout molecular dynamics simulations, thereby enhancing ligand stability within the binding pocket. In addition to the stable hydrophobic contacts, MET769 consistently emerged as a key interacting residue, forming both hydrogen-bonding and hydrophobic interactions with several derivatives, including the h, br, and methyl compounds. This residue's persistent involvement across different complexes suggests its crucial role in stabilizing ligand binding within the EGFR active site and enhancing overall complex stability. This observation is further supported by Abdelgawad et al. (2016), who found that the N-1 atom of their pyrazolo[3,4-d]pyrimidine derivatives formed a hydrogen bond with Met769, effectively anchoring the ligand within the EGFR active site. The recurrence of this residue as a critical interaction hotspot across various scaffolds underscores its pivotal role in stabilizing ligand binding and maintaining complex integrity within the EGFR kinase domain.

The remaining derivatives, hydrogen (H), chlorine (Cl), fluorine (F), and p-methoxy, are presented in Supplementary Fig. S3, where the potential of p-methoxy as a hydrogen bond acceptor is undermined by its suboptimal orientation, leading to increased solvent exposure and weaker interactions. This comparative analysis highlights the superior binding profiles of methyl, bromo, and nitro, driven by their optimized hydrophobic and polar interactions, positioning them as promising leads for further development. For nitro, moderate flexibility did not compromise binding due to strong residue interactions, whereas h remained highly stable but engaged in comparatively weaker interactions, resulting in lower binding energy. At the same time, the supplementary data provides context for refining the less effective derivatives (Grosdidier et al. 2011).

Although single 200 ns MD trajectories were sufficient for the exploratory scope of this study, we acknowledge that full convergence analyses, including RMSD stability plots as well as independent MD replicates, are important for strengthening reliability. These extended analyses will be incorporated in future work as part of a more comprehensive validation.

Toxicological profile

Pharmacokinetic and toxicological data are crucial to the rational design of effective medications, providing essential insights into the safety and viability of lead compounds in pharmaceutical research (Taft et al.). In this study, the toxicity profiles of the lead compounds were carefully evaluated in conjunction with their binding and dynamic behaviour. Table 4 below represents toxicity predictions of Olmutinib and the thiazine lead compounds. Genetic toxicology tests, such as the AMES assay, are commonly used to classify compounds as mutagenic, non-mutagenic, or inconclusive. Depending on the intended application and regulatory requirements, a mutagenic result can restrict further development, limit the compound's use, or prevent its commercialization (Zeiger 2023). In silico AMES predictions using pkCSM indicated that Olmutinib is non-mutagenic, suggesting a low risk of genetic toxicity (Suharsanti 2025; Pires et al. 2015).

Table 4.

Toxicological data of the lead compounds

Compounds AMES toxicity hERG I inhibitor hERG II inhibitor Hepatotoxicity Skin Sensitisation
Olmutinib No No Yes Yes No
Methyl Yes No Yes No No
Nitro Yes No Yes No No
Br Yes No Yes No No
Open in a new tab

In contrast, the thiazine derivatives, methyl, nitro, and br were predicted to be AMES-positive, indicating a potential risk of mutagenicity. Positive AMES results suggest that the compounds may induce point mutations, which could lead to genomic instability in these systems.

These findings do not invalidate the computational binding affinity results, but rather underscore the need for structural refinement before progressing toward drug-likeness. To mitigate this risk, several medicinal chemistry strategies can be explored, including substituting electron-withdrawing groups that maintain activity or applying prodrug approaches to reduce direct genotoxic interactions. Based on the current predictions, the methyl and bromo derivatives may represent more suitable starting points than the nitro derivative, provided mutagenicity can be successfully minimized. In addition, experimental validation, such as Ames assays and micronucleus tests, will be crucial in future work to confirm these computational alerts and guide the safe optimization of compounds.

Drug-induced hepatotoxicity encompasses both acute and chronic liver injury and remains a significant challenge in drug development, as no universally accepted primary screening methods exist for early identification of hepatotoxic compounds (Odhiambo et al. 2025). In this study, olmutinib was predicted to be hepatotoxic, whereas the thiazine derivatives were predicted to be non-hepatotoxic, suggesting a potential advantage in liver safety.

Cardiotoxicity, often assessed by hERG channel inhibition, is another key long-term safety concern, as it can lead to QT prolongation and arrhythmias (Walker et al. 1999). All compounds, including olmutinib, were predicted to inhibit hERG-II but not hERG-I, indicating a moderate risk of cardiotoxicity that should be monitored through in vitro hERG assays and in vivo cardiac studies. Future studies should include in vitro hERG patch-clamp assays to confirm these findings. Additionally, medicinal chemistry strategies, such as modifying substituents that may interact with the hERG channel, can be employed to reduce liability while retaining EGFR inhibitory activity. These steps will ensure that the selected lead compounds, particularly those with methyl and bromo substitutions, are safer for further development.

All compounds were predicted to be non-sensitizing to the skin, suggesting a low risk of dermatological toxicity. While the thiazine compounds addressed some of olmutinib's toxicity challenges, their potential mutagenicity poses a significant concern that would need to be addressed in further development.

Conclusion

This study harnessed computational methodologies to assess the potential of nine thiazine-based derivatives as inhibitors of the epidermal growth factor receptor (EGFR) in breast cancer. Molecular docking and MM-GBSA binding energy calculations identified methyl, nitro, and bromine derivatives as standout performers, exhibiting superior binding affinities and more favourable interaction energies compared to the reference compound olmutinib. ADMET profiling further underscored the pharmacokinetic strengths of these thiazine compounds, revealing excellent oral absorption, favourable solubility, minimal central nervous system penetration, and full compliance with Lipinski's Rule of Five and Jorgensen's Rule of Three.

Molecular dynamics simulations provided robust evidence of the stability and compactness of the EGFR-ligand complexes, with methyl and bromine derivatives demonstrating the lowest RMSD, RMSF, SASA, and rGyr values, indicative of highly stable and deeply embedded binding configurations. In contrast, olmutinib and the p-methoxy derivative exhibited reduced stability and more exposed conformations, highlighting their relative limitations. Toxicological evaluation revealed that the lead compounds mitigated liver toxicity and skin sensitization risks, while retaining hERG-II inhibition potential and exhibiting potential mutagenicity concerns. Collectively, these results position methyl and bromine as the most promising candidates for lead optimization, leveraging their potent binding affinities, advantageous pharmacokinetic profiles, and stable dynamic behaviour. Thiazine derivatives thus emerge as a compelling scaffold for developing next-generation EGFR inhibitors.

EGFR can develop point mutations that render it resistant to existing tyrosine kinase inhibitors, particularly T790M, L858R, and exon 19 deletions, which alter the ATP-binding site and reduce inhibitor affinity (Cross et al. 2014; Wang et al. 2016). A study conducted by Teng et al. (2011) reported EGFR mutations, specifically exon 19 deletions and exon 21 missense (L858R) mutations, in 11.8% of Triple Negative Breast Cancer (TNBC) samples analysed. All mutations were heterozygous deletions, suggesting they are likely dominant and may contribute to tumour development.

In this study, we focused on wild-type EGFR to establish a baseline understanding of thiazine–EGFR interactions. Future work will expand to mutant EGFR forms, using selectivity docking and molecular dynamics simulations to assess whether thiazine compounds maintain effective binding and stability in the presence of resistance mutations. To complement these computational studies, experimental validation will be conducted, including in vitro EGFR kinase inhibition assays, cell viability/proliferation studies in EGFR-overexpressing breast cancer cell lines, Ames test for mutagenicity, hERG patch-clamp assays to assess cardiotoxicity, and microsomal stability studies to evaluate metabolic stability. These experiments will provide critical confirmation of the bioactivity, safety, and pharmacokinetic properties of the prioritized methyl and bromine-substituted thiazine derivatives, guiding further optimization of thiazine-based EGFR inhibitors.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (859KB, docx)

Acknowledgements

The authors thank Ms Andrea Josiah for administrative support. We would also like to extend our gratitude to the Computational Chemistry group, particularly Ms. Blessing Masora and Mr. Vitalis Mbayo, for their continued support. Additionally, we acknowledge the University of Johannesburg and the Department of Chemical Sciences for providing the platform and resources that enabled this research.

Authors contributions

L.M.M., K.K.G., and V.A.O. contributed to the conceptualization and methodology. L.M.M. was responsible for investigation, data curation, formal analysis, visualization, and writing of the original draft. K.K.G. and V.A.O. contributed to supervision, validation, writing, review, and editing. P.P.G. was involved in project administration, supervision, and funding acquisition.

Funding

Open access funding provided by University of Johannesburg. This research was funded by the Mining Qualification Authority (MQA) at the University of Johannesburg.

Data availability

All data generated during this study are provided within the manuscript and the supplementary data. No additional raw datasets are available beyond what is presented.

Declarations

Competing Interests

The authors declare that they have no competing financial interests or personal relationships that could influence the work reported in this paper.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. Abbasi M, Mansourian M, Oskouie AA et al (2024) In-silico study MM/GBSA binding free energy and molecular dynamics simulation of some designed remdesivir derivatives as the inhibitory potential of SARS-CoV-2 main protease. Res Pharm Sci 19:29–41. 10.4103/1735-5362.394818 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Abdelgawad MA, Bakr RB, Alkhoja OA, Mohamed WR (2016) Design, synthesis and antitumor activity of novel pyrazolo [3,4-d] pyrimidine derivatives as EGFR-TK inhibitors. Bioorg Chem 66:88–96. 10.1016/j.bioorg.2016.03.010 [DOI] [PubMed] [Google Scholar]
  3. Adcock SA, McCammon JA (2006) Molecular dynamics: survey of methods for simulating the activity of proteins. Chem Rev 106:1589–1615. 10.1021/cr040426m [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Agarwal SM, Nandekar P, Saini R (2022) Computational identification of natural product inhibitors against EGFR double mutant (T790M/L858R) by integrating ADMET, machine learning, molecular docking and a dynamics approach. RSC Adv 12(26):16779–16789. 10.1039/D2RA00373B [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Ali RMH, Al-Hamashi AA (2024) Molecular docking, ADMET, molecular dynamic simulation, synthesis, and preliminary antiproliferative study of 1,2,4-thiadiazole derivatives as possible histone deacetylase inhibitors. Trop J Pharm Res 23:1069–1076. 10.4314/tjpr.v23i7.4 [Google Scholar]
  6. Alzain AA, Elbadwi FA, Osman W et al (2025) Broadening the scope of bacteria DNA gyrase B inhibitors by marine compounds: insights form molecular docking, MM-GBSA, molecular dynamics simulations and ADMET predictions. Sci Afr. 10.1016/j.sciaf.2025.e02658 [Google Scholar]
  7. Anderson AC (2003) The process of structure-based drug design. Chem Biol 10:787–797. 10.1016/j.chembiol.2003.09.002 [DOI] [PubMed] [Google Scholar]
  8. Baby K, Maity S, Mehta CH et al (2020) Targeting SARS-CoV-2 RNA-dependent RNA polymerase: an in silico drug repurposing for COVID-19. F1000Res. 10.12688/f1000research.26359.1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Badshah SL, Naeem A (2016) Bioactive thiazine and benzothiazine derivatives: green synthesis methods and their medicinal importance. Molecules 21(8):1054. 10.3390/molecules21081054 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bakan A, Meireles LM, Bahar I (2011) ProDy: protein dynamics inferred from theory and experiments. Bioinformatics 27:1575–1577. 10.1093/bioinformatics/btr168 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Balkrishna A, Devi K, Arya V, Dabas A (2024) Exploring therapeutic potentials of Tridax procumbens for glioma: an in-silico approach employed network pharmacology, molecular docking, and molecular dynamics simulation. S Afr J Bot 175:122–141 [Google Scholar]
  12. Banks JL, Beard HS, Cao Y et al (2005) Integrated modeling program, applied chemical theory (IMPACT). J Comput Chem 26:1752–1780. 10.1002/jcc.20292 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Baselga J, Arteaga CL (2005) Critical update and emerging trends in epidermal growth factor receptor targeting in cancer. J Clin Oncol 23:2445–2459. 10.1200/JCO.2005.11.890 [DOI] [PubMed] [Google Scholar]
  14. Berasain C, Perugorria MJ, Latasa MU et al (2009) The epidermal growth factor receptor: a link between inflammation and liver cancer. Exp Biol Med 234:713–725 [DOI] [PubMed] [Google Scholar]
  15. Bharadwaj KK, Ahmad I, Pati S et al (2024) Screening of Phytocompounds for Identification of Prospective Histone Deacetylase 1 (HDAC1) Inhibitor: an in silico molecular docking, molecular dynamics simulation, and MM-GBSA approach. Appl Biochem Biotechnol 196:3747–3764. 10.1007/s12010-023-04731-3 [DOI] [PubMed] [Google Scholar]
  16. Bhargavi M, Gayathri Bhargavi S, Sripada E (2025) Computational approach for HDAC1 predicting protein-ligand interactions for cancer through homology modelling, virtual screening and molecular docking. In: Convergence of technology & biology - transforming life sciences. Shanlax Publications, pp 8–21
  17. Chiodi D, Ishihara Y (2024) The role of the methoxy group in approved drugs. Eur J Med Chem 273:116364. 10.1016/j.ejmech.2024.116364 [DOI] [PubMed] [Google Scholar]
  18. Cross DA et al (2014) AZD9291, an irreversible EGFR TKI, overcomes T790M-mediated resistance to EGFR inhibitors in lung cancer. Cancer Discov 4(9):1046–1061. 10.1158/2159-8290.CD-14-0337 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Dulsat J, López-Nieto B, Estrada-Tejedor R, Borrell JI (2023) Evaluation of free online ADMET tools for academic or small biotech environments. Molecules 28:65 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Elsebaie HA, El-Moselhy TF, El-Bastawissy EA et al (2024) Development of new thieno[2,3-d]pyrimidines as dual EGFR and STAT3 inhibitors endowed with anti-cancer and pro-apoptotic activities. Bioorg Chem. 10.1016/j.bioorg.2024.107101 [DOI] [PubMed] [Google Scholar]
  21. Emami L et al (2022) Synthesis, biological evaluation, and computational studies of some novel quinazoline derivatives as anti-cancer agents. BMC Chem 16(1):100 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Fereidounpour P, Steinmann C, Larsen KL (2024) Prediction of the free energy of binding for cyclodextrin–steroid complexes: phase solubility and molecular dynamics studies. J Incl Phenom Macrocycl Chem 104:535–546. 10.1007/s10847-024-01255-z [Google Scholar]
  23. Gheidari D, Mehrdad M, Hoseini F (2024) Virtual screening, molecular docking, MD simulation studies, DFT calculations, ADMET, and drug likeness of Diaza-adamantane as potential MAPKERK inhibitors. Front Pharmacol. 10.3389/fphar.2024.1360226 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Goodsell DS, Olson AJ (1990) Automated docking of substrates to proteins by simulated annealing. Proteins Struct Funct Bioinform 8:195–202. 10.1002/prot.340080302 [DOI] [PubMed] [Google Scholar]
  25. Grosdidier A, Zoete V, Michielin O (2011) SwissDock, a protein-small molecule docking web service based on EADock DSS. Nucleic Acids Res 39:W270–W277. 10.1093/nar/gkr366 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Gunasekaran V, Dhakshinamurthy SS (2024) Computational insights into the interaction of pinostrobin with Bcl-2 family proteins: a molecular docking analysis. Asian Pac J Cancer Prevent 25:507–512. 10.31557/APJCP.2024.25.2.507 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hamad HT (2025) The anti-cancer effectiveness of some heterocyclic compounds containing sulfur atom. Results Chem 15:102182. 10.1016/j.rechem.2025.102182 [Google Scholar]
  28. Hynes NE, Lane HA (2005) ERBB receptors and cancer: the complexity of targeted inhibitors. Nat Rev Cancer 5:341–354. 10.1038/nrc1609 [DOI] [PubMed] [Google Scholar]
  29. Ismail RS, Issa AH, Almayah AA (2019) Immunohistochemistry of Detection HER2 on Breast Cancer in Basra City
  30. Kalyaanamoorthy S, Chen Y-PP (2011) Structure-based drug design to augment hit discovery. Drug Discov Today 16:831–839. 10.1016/j.drudis.2011.07.006 [DOI] [PubMed] [Google Scholar]
  31. Kendall RA, Dunning TH, Harrison RJ (1992) Electron affinities of the first-row atoms revisited. Systematic basis sets and wave functions. J Chem Phys 96:6796–6806. 10.1063/1.462569 [Google Scholar]
  32. Kesharwani RK, Vishwakarma VK, Keservani RK, Singh P, Katiyar N, Tripathi S (2020) Role of ADMET tools in current scenario: application and limitations. In: Singh DB (ed) Computer-aided drug design. Springer, Singapore. 10.1007/978-981-15-6815-2_4 [Google Scholar]
  33. Kim S, Chen J, Cheng T et al (2023) PubChem 2023 update. Nucleic Acids Res 51:D1373–D1380. 10.1093/nar/gkac956 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kitchen DB, Decornez H, Furr JR, Bajorath J (2004) Docking and scoring in virtual screening for drug discovery: methods and applications. Nat Rev Drug Discov 3:935–949. 10.1038/nrd1549 [DOI] [PubMed] [Google Scholar]
  35. Kumar S, Ayyannan SR (2023) Identification of new small molecule monoamine oxidase-B inhibitors through pharmacophore-based virtual screening, molecular docking and molecular dynamics simulation studies. J Biomol Struct Dyn 41(14):6789–6810. 10.1080/07391102.2022.2112082 [DOI] [PubMed] [Google Scholar]
  36. Kumar S, Arora A, Sapra S, Kumar R, Singh BK, Singh SK (2024) Recent advances in the synthesis and utility of thiazoline and its derivatives. RSC Adv 14(2):902–953 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Llop-Peiró A, Macip G, Garcia-Vallvé S, Pujadas G (2024) Are protein–ligand docking programs good enough to predict experimental poses of noncovalent ligands bound to the SARS-CoV-2 main protease? Drug Discov Today 29:65 [DOI] [PubMed] [Google Scholar]
  38. Masuda H, Zhang D, Bartholomeusz C et al (2012) Role of epidermal growth factor receptor in breast cancer. Breast Cancer Res Treat 136:331–345. 10.1007/s10549-012-2289-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mbayo V, Govender PP, Marondedze EF, Govender KK (2025) Quantitative structure-activity relationship, activity prediction, and molecular dynamics of non-nucleotide reverse transcriptase inhibitors. J Visual Exp. 10.3791/67457 [DOI] [PubMed] [Google Scholar]
  40. Meng X-Y, Zhang H-X, Mezei M, Cui M (2011) Molecular docking: a powerful approach for structure-based drug discovery. Curr Comput Aided-Drug des 7:146–157. 10.2174/157340911795677602 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Moussaoui M et al (2024) Design and optimization of quinazoline derivatives as potent EGFR inhibitors for lung cancer treatment: a comprehensive QSAR, ADMET, and molecular modeling investigation. ACS Omega 9(46):45842–45857. 10.1021/acsomega.4c04639 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Mukherjee S, Balius TE, Rizzo RC (2010) Docking validation resources: protein family and ligand flexibility experiments. J Chem Inf Model 50:1986–2000. 10.1021/ci1001982 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Naidoo V, Achilonu I, Mirza S et al (2025) Computational modelling of tunicamycin C interaction with potential protein targets: perspectives from inverse docking with molecular dynamic simulation. Curr Issues Mol Biol. 10.3390/cimb47050339 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. National Cancer Institute (2023) Cancer treatment side effects
  45. Obakachi VA, Govender KK, Govender PP (2025) Targeting Acute Myeloid Leukemia with 1,2,4-triazolo[4,3-b]pyridazine derivatives: a molecular docking, dynamics, and ADMET approach. Silico Pharmacol 13:135. 10.1007/s40203-025-00418-1 [DOI] [PMC free article] [PubMed]
  46. Odhiambo DO, Omosa LK, Njagi EC, Kithure JG, Wekesa EN (2025) In-silico pharmacokinetics ADME/Tox analysis of phytochemicals from genus Dracaena for their therapeutic potential. Sci Afr 8:e02796. 10.1016/j.sciaf.2025.e02796 [Google Scholar]
  47. Ottu PO, Babalola OO, Oluwamodupe C et al (2025) Investigation of Aframomum melegueta compounds as ERK5 inhibitor related to breast cancer via molecular docking and dynamic simulation. In Silico Pharmacol 13:18. 10.1007/s40203-025-00304-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Palanivel S, Yli-Harja O, Kandhavelu M (2021) Molecular interaction study of novel indoline derivatives with EGFR-kinase domain using multiple computational analysis. J Biomol Struct Dyn 40(16):7545–7554. 10.1080/07391102.2021.1900917 [DOI] [PubMed] [Google Scholar]
  49. Pan X, Wang H, Zhang Y et al (2022) AA-Score: a new scoring function based on amino acid-specific interaction for molecular docking. J Chem Inf Model 62:2499–2509. 10.1021/acs.jcim.1c01537 [DOI] [PubMed] [Google Scholar]
  50. Pires DE, Blundell TL, Ascher DB (2015) pkCSM: predicting small-molecule pharmacokinetic and toxicity properties using graph-based signatures. J Med Chem 58(9):4066–4072. 10.1021/acs.jmedchem.5b00104 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. PKCMS PKCMS: Pharmacokinetics modeling & simulation platform. https://www.pkcms.com/. Accessed 20 Jun 2025
  52. RCSB Protein Data Bank (2024) RCSB PDB: Homepage. https://www.rcsb.org. Accessed 20 Jun 2025
  53. Schrödinger Release 2023-2 (2023a) QikProp. Schrödinger, LLC
  54. Schrödinger Release 2023-2 (2023b) Desmond molecular dynamics system. maestro-desmond interoperability tools
  55. Sanphanya K, Wattanapitayakul SK, Phowichit S et al (2013) Novel VEGFR-2 kinase inhibitors identified by the back-to-front approach. Bioorg Med Chem Lett 23:2962–2967. 10.1016/j.bmcl.2013.03.042 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Schrödinger L (2023a) Maestro 13.6 User Manual. Schrödinger, LLC
  57. Schrödinger L (2023b) QikProp User Manual, version 13.6. Schrödinger, LLC
  58. Sethi A, Joshi K, Sasikala K, Alvala M (2020) Molecular docking in modern drug discovery: principles and recent applications. IntechOpen. 10.5772/intechopen.85991 [Google Scholar]
  59. Sharma PK, Makkar R (2016) A review: thiazines derivatives treated as potential antimicrobial agents. Asian J Pharm Clin Res 10:43. 10.22159/ajpcr.2017.v10i1.115467 [Google Scholar]
  60. Sharma SV, Bell DW, Settleman J, Haber DA (2007) Epidermal growth factor receptor mutations in lung cancer. Nat Rev Cancer 7:169–181. 10.1038/nrc2088 [DOI] [PubMed] [Google Scholar]
  61. Shi K et al (2022) Emerging strategies to overcome resistance to third-generation EGFR inhibitors. J Hematol Oncol. 10.1186/s13045-022-01311-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Shri SR, Nayak Y, Ranganath Pai S (2024) Molecular docking studies and molecular dynamic simulation analysis: to identify novel ATP-competitive inhibition of Glycogen synthase kinase-3β for Alzheimer’s disease. F1000Res 13:773. 10.12688/f1000research.145391.1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Suharsanti R (2025) Molecular docking of lipase inhibitory activities, pharmacokinetics and toxicity prediction of chemical constituents from curcuma aeruginosa Roxb rhizome. Repository Stifar
  64. Sung H, Ferlay J, Siegel RL et al (2021) Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 71:209–249. 10.3322/caac.21660 [DOI] [PubMed] [Google Scholar]
  65. Taft CA, Ricardo S, Editors L Engineering Materials
  66. Tebbutt N, Pedersen MW, Johns TG (2013) Targeting the ERBB family in cancer: couples therapy. Nat Rev Cancer 13:663–673. 10.1038/nrc3559 [DOI] [PubMed] [Google Scholar]
  67. Teng YHF et al (2011) Mutations in the epidermal growth factor receptor (EGFR) gene in triple negative breast cancer: possible implications for targeted therapy. Breast Cancer Res 13:R35. 10.1186/bcr2857 [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. The Halogens: Discovery, Occurrence, and Biochemistry of the Free Elements
  69. Tilak BD (1960) Thiazine and its derivatives: a review of their chemistry and biological activities. In: Proceedings of the Indian Academy of Sciences
  70. Türkmenoğlu B (2022) Investigation of novel compounds via in silico approaches of EGFR inhibitors as anti-cancer agents. J Indian Chem Soc 99(8):100601 [Google Scholar]
  71. Van Der Spoel D, Lindahl E, Hess B et al (2005) GROMACS: fast, flexible, and free. J Comput Chem 26:1701–1718. 10.1002/jcc.20291 [DOI] [PubMed] [Google Scholar]
  72. Waks AG, Winer EP (2019) Breast cancer treatment. JAMA 321:288. 10.1001/jama.2018.19323 [DOI] [PubMed] [Google Scholar]
  73. Walker BD, Singleton CB, Bursill JA, Wyse KR, Valenzuela SM, Qiu MR, Breit SN, Campbell TJ (1999) Inhibition of the human ether-a-go-go-related gene (HERG) potassium channel by cisapride: affinity for open and inactivated states. Br J Pharmacol 128(2):444–450. 10.1038/sj.bjp.0702774 [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Wang S, Cang S, Liu D (2016) Third-generation inhibitors targeting EGFR T790M mutation in advanced non-small cell lung cancer. J Hematol Oncol 9:34. 10.1186/s13045-016-0268-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Wang G, Liu W, Fan M, He M, Li Y, Peng Z (2021) Design, synthesis and biological evaluation of novel thiazole–naphthalene derivatives as potential anti-cancer agents and tubulin polymerisation inhibitors. J Enzyme Inhib Med Chem 36(1):1694–1702 [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Wani MA, Banerjee A, Garg P (2025) Computer-aided drug design approaches for the identification of potent inhibitors targeting elongation factor G of Mycobacterium tuberculosis. J Mol Graph Model. 10.1016/j.jmgm.2025.108954 [DOI] [PubMed] [Google Scholar]
  77. World Health Organization (2022) Cancer fact sheet
  78. Yarden Y, Sliwkowski MX (2001) Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2:127–137. 10.1038/35052073 [DOI] [PubMed] [Google Scholar]
  79. Yazdani B, Sirous H, Enguita FJ et al (2024) Discovery of novel direct small-molecule inhibitors targeting HIF-2α using structure-based virtual screening, molecular dynamics simulation, and MM-GBSA calculations. Mol Divers 28:1203–1224. 10.1007/s11030-023-10650-6 [DOI] [PubMed] [Google Scholar]
  80. Zare S et al (2023) 6-bromoquinazoline derivatives as potent anti-cancer agents: synthesis, cytotoxic evaluation, and computational studies. Chem Biodivers 20(7):e202201245 [DOI] [PubMed] [Google Scholar]
  81. Zeiger E (2023) Determination of a positive response in the Ames Salmonella mutagenicity assay. Environ Mol Mutagen 64(4):250–258. 10.1002/em.22538 [DOI] [PubMed] [Google Scholar]
  82. Zev S, Raz K, Schwartz R et al (2021) Benchmarking the ability of common docking programs to correctly reproduce and score binding modes in SARS-CoV-2 protease Mpro. J Chem Inf Model 61:2957–2966. 10.1021/acs.jcim.1c00263 [DOI] [PubMed] [Google Scholar]
  83. Zhao Y, Truhlar DG (2008) The M06 suite of density functionals for main group thermochemistry, thermochemical kinetics, noncovalent interactions, excited states, and transition elements: two new functionals and systematic testing of four M06-class functionals and 12 other functionals. Theor Chem Acc 120:215–241. 10.1007/s00214-007-0310-x [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (859KB, docx)

Data Availability Statement

All data generated during this study are provided within the manuscript and the supplementary data. No additional raw datasets are available beyond what is presented.

Articles from In Silico Pharmacology are provided here courtesy of Springer

RESOURCES