https://orcid.org/0009-0000-2565-6573
Figures
Abstract
Programmed death-ligand 1 (PD-L1), a transmembrane protein, is associated with the regulation of immune system. It frequently has overexpression in various cancers, allowing tumor cells to avoid immune detection. PD-L1 inhibition has risen as a potential strategy in the field of therapeutic immunology for cancer. In the current study, structure-based virtual screening of drug libraries was conducted and then the screened hits were docked to the active residues of PD-L1 to select the optimal binding poses. The top ten compounds with binding affinities ranging from -10.734 to -10.398 kcal/mol were selected for further analysis. The ADMET analysis of selected compounds showed the compounds meet the criteria of ADMET properties. Further, the conformational changes and binding stability of the top two compounds was analyzed by conducting 200 ns simulation and it was observed that the hits did not exert conformational changes to the protein structure. All the results suggest that the chosen hits can be considered as lead compounds for the inhibition of biological activity of PD-L1 in in vitro studies.
Citation: Alanzi A, Moussa AY, Mothana RA, Abbas M, Ali I (2024) In silico exploration of PD-L1 binding compounds: Structure-based virtual screening, molecular docking, and MD simulation. PLoS ONE 19(8): e0306804. https://doi.org/10.1371/journal.pone.0306804
Editor: Sheikh Arslan Sehgal, The Islamia University of Bahawalpur Pakistan, PAKISTAN
Received: March 18, 2024; Accepted: June 23, 2024; Published: August 9, 2024
Copyright: © 2024 Alanzi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data is incorporated in the manuscript.
Funding: The author(s) received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
A transmembrane protein called Programmed death-ligand 1 (PD-L1) is affiliated with immune regulation. It is associated with the B7 family of cell surface proteins and is found on various cell type’s surfaces, comprising immune cells and some cancer cells. The receptor for PD-L1, known as programmed cell death protein 1 (PD-1), is found on the T cells’ surface and other cells of immune system. The activity of T cell is controlled by this interaction and prevents overreactions from the immune system, acting as a crucial immune checkpoint [1–3].
The primary function of PD-L1 is to suppress the immune system, specifically the activity of cytotoxic T cells. Upon the occurrence of PD-L1 binding to PD-1 on the surfaces of T cells, it prevents them from activating or functioning. This mechanism is essential for sustaining immune tolerance and inhibiting the immune system to attack healthy cells. Certain tumors, however, take advantage of this pathway by overexpressing PD-L1, allowing them to avoid immune detection and destruction [4, 5].
The interaction between PD-L1 and PD-1 is targeted for therapeutic approaches in cancer immunotherapy. In recent years, the checkpoint inhibitors of immune system which block the PD-L1/PD-1 interaction have emerged as a novel approach to treat the cancer [6–8]. By inhibiting this interaction, these inhibitors enable the immune system to identify and combat cancerous cells [8, 9]. Clinical research has shown that inhibiting the interactions between PD-1/PD-L1 can increase the immune response mediated by T cells against cancer, produce prolonged clinical effects, and the patient survival rate increases [10, 11].
The FDA has approved monoclonal antibodies (mAbs) that target either PD-L1 (e.g., Avelumab, Durvalumab, and Atezolizumab) or PD-1 (e.g., Nivolumab, Cemiplimab, and Pembrolizumab) to treat a number of cancers [12–14]. The clinical application of immune checkpoint inhibitors based on antibodies is restricted due to immunogenicity, immunological-related side effects, and high costs, even though these mAbs show promising therapeutic efficacy in clinical trials [12, 15]. Furthermore, the large size of these mAbs results in limited permeability in the tumor tissues. They can cause serious side effects because of their comparatively long half-life, which makes drug elimination more difficult [16, 17].
As an alternative, small molecule inhibitors might have advantageous oral bioavailability and tumor penetration. Additionally, compared to monoclonal antibodies (mAbs), which have garnered a lot of interest in the pharmaceutical industry, small molecule inhibitors may result in few side effects, be simpler to self-administer, possess a shorter biological half-life, and cost effective. However, majority of low molecular weight inhibitor of PD-1/PD-L1 are still in the phase of their early development, emphasizing preclinical research [17, 18]. Hence, this study has been designed to identify potential inhibitors against PD-L1 protein utilizing pharmacophore modeling, virtual screening, molecular docking, and molecular dynamics simulation. The flow chart of the study is shown in Fig 1.
2. Methodology
2.1. Virtual screening
For the selection of a limited number of compounds capable of interacting with the protein’s binding pocket, structure-based virtual screening proves to be a reliable method. The 3D grid parameters are essential for guiding the software in defining the specific search space for virtual screening.
2.2. Parametrization for 3D grid
A 3D grid was generated by defining the space surrounding the binding pocket of PD-L1. This grid was employed for structure-based virtual screening, and customized for the calculation of X, Y, and Z coordinates by using AutoDock tools [19].
2.3. Virtual screening using MTiOpenScreen
Virtual screening for was performed using MTiOpenScreen, that is an efficient web server for structure based virtual screening [20]. MTiOpenScreen provides valuable libraries of classified compounds including a collection of compounds derived from natural products (NP-lib), Approved drugs available for purchase (Drugs-lib), a collection of food constituent compounds (FOOD-lib), and a diverse drug-like library (Diverse-lib). The diverse library is an assembly of 12 chemical libraries from PubChem BioAssay Database which is prepared by FAFDrugs3 web-server [21]. This library contains a total of 99,288 drug like molecules. In this study, we employed virtual screening of diverse libraries. The MTiOpenScreen server supports the automation of Autodock Vina for virtual screening [22] and 100 compounds are screened based on the binding affinities.
2.4. Molecular docking
In Maestro, the protein preparation wizard was employed for the preparation of the crystal structure of PD-L1 (PDB ID: 5N2F) [23]. The preprocessing of the receptor has been performed by addition of hydrogens, eliminating additional water molecules, incorporating appropriate charges, and fixing the side chain residues. While generating the tautomeric variations at pH 7.0 by using PROPKA, the elimination of unnecessary ligands and chains was performed [24]. The optimization and minimization of the receptor’s structure were then performed using the OPLS_2005 forcefield [25]. Following the protein preparation, the construction of the 3D grid involved selecting the natural substrate for site-specific docking. To mitigate the impact of non-polar regions of the receptor, the van der Waals radii of the receptor atom were set to 1.0, and the partial charge cutoff value was scaled at 0.25. X, Y, and Z coordinates have values approximately 33.109, 6.139, and 131.907 respectively. After generating the grid, LigPrep tool of Maestro was utilized to perform the ligands preparation before docking [26]. Various ionization states were generated under pH 7 conditions using Epik. The OPLS_2005 forcefield was used to produce with defined chirality. Subsequently, the docking of prepared ligand was performed to the prepared receptor by utilizing the Glide docking tool and the analysis of binding modes was done depending on the glide score.
2.5. ADMET analysis
The top ten compounds selected based on the docking scores were subjected to SwissADME webserver (http://www.swissadme.ch/index.php) to calculate the pharmacokinetic properties. For a molecule to be considered as a drug, must pass different filters i.e., Lipinski rule of five. Similarly, the toxicity profiles of the compounds were predicted by using the ProTox-II webtool (https://tox.charite.de/protox3/). The insilico toxicity prediction of the compounds reduces the cost of experimentation. The toxicity properties such as Heptatoxicity, Carcinogenicity, Immunogenicity, Mutagenicity, and Cytotoxicity were predicted and compared.
2.6. Molecular dynamics (MD) simulation
MD simulations were conducted on selected compounds using Desmond for a duration of 200 ns [27]. MD simulations were executed to assess the stability of protein-ligand complexes resulting from the docking process. The preparation of complexes for simulation included Preprocessing, optimization, and minimization. We have done the minimization of process by using the OPLS_2005 force field [25]. The system solvation was achieved by using an orthorhombic box measuring 10 Å, which included TIP3P water molecules [28]. For the systems’ charge neutralization 0.15 M NaCl was added. The system was equilibrated under the NPT ensemble, and the NVT ensemble, established at a temperature of 300K and under a pressure of 1 atm. The systems were allowed to relax, before proceeding to simulations. Finally, employing trajectory snapshots of 40 ps, simulation was performed, allowing for subsequent examination of the results.
3. Results
3.1. 3D grid parametrization
The 3D grid box was defined around the binding pocket’s active residues of the PD-L1 protein. The 3D coordinates were determined by considering the co-crystal ligand. The intrinsic coordinates for x, y, and z were 33.109, 6.139, and 131.907 respectively. The visual representation of 3D grid box is given in Fig 2.
3.2. Virtual screening of PD-L1 in MTiOpenScreen
MTiOpenScreen was employed to perform the structure-based virtual screening of the PD-L1 protein by utilizing the diverse library containing 99,288 diverse drug-like compounds depending on the binding affinities [29]. 100 compounds were screened against the PD-L1 in total.
3.3. Molecular docking
The screened compounds generated during virtual screening underwent docking to the prepared PD-L1 receptor to predict the binding affinities by utilizing the default precision mode of glide tool. Based on the binding affinities, the top ten compounds were selected for further analysis (Table 1). The selected compounds had their binding affinities ranging from -10.734 to -10.398 kcal/mol, indicating the probability of inhibiting the function of PD-L1 protein of selected compounds.
3.4. Molecular interactions analysis
The interactions analysis of the selected hits with the binding pocket of PD-L1 receptor was performed using the Discover Studio client tool [30]. The observed interactions mainly involved: conventional hydrogen bond, carbon hydrogen bond, van der Waal interactions, Pi-Sulfur, Amide Pi-Stacked, Halogen, and Alkyl interactions. These interactions contribute to determining the binding affinities and docking scores for each of the top candidate compounds. Notably, the establishment of intermolecular hydrogen bonds formed between the ligand and the amino acid within the active sites significantly influence the overall strength of the resulting complex. Consequently, these interactions consistently enhance the docking results [31]. Compound-103051331 formed one conventional hydrogen bond with Ile116B, two Pi-Sigma interactions with Ala121A, Ala121B, two Pi-Sulfur interactions with Met115A, Met115B, and four alkyl interactions with Tyr56A, Tyr123B, Ile54B, Tyr56B (Fig 3A). Compound-57264996 formed two conventional hydrogen bonds with Ile116A, Asp122A, two Pi-Sigma interactions with Ala121A, Ala121B, two Pi-Sulfur interactions with Met115A, Met115B, one halogen interaction with Ile116B, and four alkyl interactions with Tyr56A, Ile54B, Val68B, Tyr56B (Fig 3B). Similarly, Compound-49642753 made two Pi-Sigma interactions with Ala121A, Met115B, two halogen interactions with Asp122B, Ala121B, and four alkyl interactions with Tyr56B, Ile54B, Tyr56A, Met115A (Fig 3C). Lastly, Compound-24362221 made two conventional hydrogen bonds with Ser117B, Gln66B, three Pi-Sigma interactions with Ala121A, Ala121B, Met115B, one Pi-Sulfur interaction with Met115A, and one alkyl interaction with Tyr56B (Fig 3D). The molecular interactions for other compounds are provided in Table 2.
A) Compound-103051331, B) Compound-57264996, C) Compound-49642753, and D) Compound-24362221. Green spheres are showing hydrogen bonds, Pi-Sigma interactions with purple spheres, magenta spheres are indicating hydrophobic interactions, orange spheres are depicting salt bridges, and cyan spheres are showing halogen interaction.
3.5. ADMET analysis
The ADME properties of the selected compounds were predicted by using the SwissADME webserver. The drug-likeness of the compounds was evaluated by Lipinski rule of five i.e., molecular wright < 500, hydrogen bond donors < 5, hydrogen bond acceptors < 10, and cLogP < 5. From the analysis, it was observed that all the compounds followed the Lipinski rule and did not show any violation (Table 3). Further the polar surface area of the compound was calculated which is important for intestinal absorption and blood brain barrier penetration. The value of TPSA more than 140 shows the poor cell membrane permeability, however, all the compounds showed the TPSA values less than 120. The highest TPSA value was 118.90 showed by Compound-24362221. Similarly, the toxicity profiles of the compounds were predicted by ProTox-II server. It can be observed that most of the compounds showed toxicity class 4 while Compound-49672076 showed toxicity class 5 (Table 4). Additionally, most of the compounds did not show active profiles for heptatoxicity, carcinogenicity, immunotoxicity, mutagenicity, and cytotoxicity. However, Compound-24285320 showed active profiles for the toxicities. After ADMET analysis, the top two compounds were selected for the stability analysis.
3.6. Binding pose analysis
Following the analysis of molecular interactions, the potential binding poses of the two selected compounds were analysed by aligning them with the co-crystal ligands. The analysis revealed that the docked compounds exhibited a precise alignment with the co-crystal ligand, demonstrating a similar binding mode (Fig 4). Thus, the stability of the selected compounds’ potential binding modes was evaluated through MD Simulation studies.
(A) Compound-49642753 (Cyan sticks), (B) and Compound-857389 (Magenta sticks).
3.7 MD simulation
3.7.1 RMSD.
To assess the stability of the protein-ligand complexes, PD-L1 receptor underwent 200 ns molecular dynamics (MD) simulation for the complexes’ chosen binding poses. Root mean square deviation (RMSD) of Carbon alpha atoms of protein and the ligand atoms fit on the protein was calculated to reflect the overall structural changes and deviations of the complexes throughout the simulation [32]. The values of RMSD for the Compound-49642753 complex exhibited deviations ranging from 2–4 Å up to 80 ns, followed by a gradual decrease to 3 Å at 100 ns. Subsequently, these values were sustained within this range throughout the entire simulation period while the RMSD of ligand fit on the protein remained in the range of 1–1.5 Å throughout the simulation (Fig 5A). A similar trend in the RMSD of Compound-857389 was observed but here RMSD elevated to 6 Å but the deviations stopped at 80 ns and then it maintained the 5 Å range until the conclusion of simulation, while the RMSD of ligand fit remained less than 1 Å throughout the simulation (Fig 5B).
(A) Compound-49642753, (B) Compound-857389.
3.7.2 RMSF.
The values for Root mean square fluctuations (RMSF) were acquired to investigate the dynamic properties exhibited by the proteins in association with the ligands [33]. The values for RMSF depict the information regarding the variations in flexibility and mobility of residues of the protein during the simulation period. The results indicated that most protein residues underwent only minor changes, with fluctuations under 2 Å, implying stability and low flexibility in the ligands’ presence. However, the loop regions of the protein, specifically residues at N- and C-terminals had demonstrated higher RMSF values, reaching around 9Å (Fig 6). The overall stability of the protein-ligand complex was evident, as most residues maintained a rigid form. The increased fluctuations in the loop regions suggest their potential role in dynamic interactions with the ligands. In summary, the RMSF analysis confirms a generally stable protein-ligand complex, characterized by limited variability in most residues but increased flexibility in certain loop sections.
(A) Compound-49642753, (B) Compound-857389.
3.7.3 Protein-ligand contacts.
In the analysis of molecular dynamics (MD) simulations, the key interactions between the protein and ligands were determined to include ionic bonds, hydrogen bonds, and hydrophobic interactions. These interactions play a vital role in maintaining stability and influencing the functional properties of the protein-ligand complex. Compound-49642753 was involved in hydrophobic interactions with the residues Ile54A, Tyr56A, Met115A, Ala121A, Tyr231A, Ilr54B, Tyr56B, Met115B, Ala121B, and Tyr123B (Fig 7A). In the Compound-857389 complex, the residues forming hydrophobic interactions were Ile54A, Tyr56A, Met115A, Ala121A, Tyr123A, Ile54B, Tyr56B, Met115B, Ala121B, and Tyr123B (Fig 7B). These specific hydrogen bonding interactions noted in the MD simulations underline the crucial residues contributing to the stability of the protein-ligand complexes, offering valuable insights into the essential interactions that govern their stability and binding affinity.
(A) Compound-49642753, (B) Compound-857389. The interacting residues are shown are shown as large, stacked bars. Gray shows the hydrophobic interactions, and blue shows the water bridges.
4. Discussion
The study of the role PD-L1 in immune regulation remains a crucial area of study in both cancer biology and immunology. Understanding the molecular mechanisms underlying PD-L1 expression and interaction with PD-1 has provided important insights for targeted therapies’ development that employ the immune system to combat cancer. PD-L1 inhibition is now recognized as a promising approach in cancer immunotherapy [2, 3].
Currently, several PD-L1 inhibitors are being developed [34, 35]. KN035 is the only PD-L1 with a subcutaneous formulation antibody, and it is currently in clinical trials in the United States, China, and Japan [36]. CA-170, developed by Aurigene and Curis, is the only small-molecule antagonist for PD-L1 and VISTA currently in clinical trials. The compound is currently undergoing phase 1 clinical trials in patients with mesothelioma [37]. However, there are a few issues with the clinical use of monoclonal antibodies that target PD-1/PD-L1, including low patient response rates, severe immune-related adverse reactions (poor immunogenicity), intravenous administration requirements (no oral bioavailability), drug resistance, etc., all of which encourage the synthesis of macrocycles, peptides, and small molecules [38, 39]. To meet the increasing clinical demands, novel inhibitors targeting the PD-1/PD-L1 axis, such as small molecules, peptides, or macrocycles, are urgently needed.
Small molecules are important in drug discovery as they have their ability to efficiently penetrate cells, interact with specific targets, and modulate biological processes, which provides benefits in terms of synthetic accessibility, oral bioavailability, and pharmacokinetic property optimization. Their versatility enables the development of diverse therapeutic agents for a wide range of diseases, making them indispensable in pharmaceutical research and clinical practice [40, 41]. In silico/computational methods provide a systematic approach to identifying small molecules which are having capability of binding to PD-L1, disrupting its interaction with programmed cell death protein 1 (PD-1), and reactivating the immune response against cancer cells. In silico/ computational approach significantly speeds up the drug discovery process, providing a cost-efficient and time-saving method for the identification of potential drug candidates with high precision. It reduces the need for time-consuming and resource-intensive experimental methods by providing a rational and systematic approach to selecting compounds with the highest chances of success [42–44]. Virtual screening is a key component of modern drug discovery efforts, providing a valuable method for quickly sorting through large chemical libraries and identifying possible drug candidates with significant biological activity. By quickly identifying lead compounds for additional experimental validation, this method speeds up the early stages of drug discovery and consequently accelerates the entire drug development process. The importance of virtual screening in drug discovery was demonstrated by studies that had used virtual screening for the discovery of novel anticancer flavonoids as potential HDAC2 and VEGFR-2 Inhibitors [45, 46], and novel urushiol derivatives as potential HDAC2 selective inhibitors [47].
This study focuses on computational screening of the drug library to identify potential PD-L1 inhibitors through a synergistic approach that combines virtual screening, molecular docking, and molecular dynamics (MD) simulations. The crystal structure of PD-L1 protein was obtained from PDB database (PDB ID: 5N2F) and prepared. Subsequently, A customized 3D grid box was defined around the residues of the active pocket of the PD-L1 protein. A structure-based virtual screening of the diverse library consisting of 99,288 diverse drug-like compounds was performed depending on the binding affinities, and 100 compounds were screened against the PD-L1 in total. The hit compounds identified during virtual screening were docked to the prepared PD-L1 receptor to predict binding affinities using the glide tool’s standard precision mode. This step aids in predicting the compounds’ potential efficacy in inhibiting PD-L1 enzymatic activity [48]. The top ten compounds were then proceeded for further investigation depending on their binding affinities. The selected compounds exhibited binding affinities within the range of -10.734 to -10.398 kcal/mol, suggesting that they could inhibit the function of the PD-L1 protein.
The molecular interactions of the selected hits with the binding pocket of the PD-L1 receptor were investigated. Conventional hydrogen bonds, carbon hydrogen bonds, van der Waal interactions, Pi-Sulfur, Pi-Pi Stacked, Pi-Sigma, and Alkyl interactions were the most observed interactions. These interactions are critical in determining each of the top candidate compounds’ binding affinities and docking scores.
Additionally, assessing ADMET properties is an important part of drug development. Computational tools are used to assess the pharmacokinetic and toxicological profiles of the identified PD-L1 inhibitors, providing information about their potential efficacy and safety. This step is critical for removing potentially harmful pharmacological properties early in the drug discovery process, thus saving time and resources [49]. The predicted ADMET properties of the chosen compounds were within acceptable ranges for all values [50, 51].
Following the molecular interaction analysis, the top two compounds’ plausible binding modes were investigated by aligning them on the co-crystal ligand. The docked compounds exhibited a precise alignment with the co-crystal ligand, demonstrating a similar binding pose, according to the analysis.
Molecular Dynamics (MD) simulations are utilized to explore the dynamic behavior of complexes involving PD-L1 inhibitors throughout a given period. Researchers can use this computational technique to investigate the stability and flexibility of binding interactions, which provides valuable information on structural changes that may affect the inhibitor’s efficacy [52]. These compounds sustained stability within the protein binding pocket pretending to be effective inhibitors, at the conclusion of simulation. According to these results, the chosen hit compounds exhibit the potential to act as lead compounds for inhibiting the biological activity of PD-L1.
Computational approaches depend on experimental methods to gather raw data for further analysis. The accuracy of results can be limited by data quality and the efficiency of the computational algorithms used. Ethical and regulatory concerns also arise, necessitating careful validation and review of computational predictions before clinical application [53, 54]. Hence, experimental validation is required to confirm these potential inhibitors into clinical drugs.
5. Conclusion
This article investigates the in-silico exploration of PD-L1 binding compounds using structure-based virtual screening, molecular docking, and molecular dynamics (MD) simulations. The findings of this investigation may open avenues for the design and evolution of therapeutic agents with improved efficacy and selectivity against PD-L1, addressing the unmet medical needs associated with PD-L1 -related diseases.
Acknowledgments
Authors extend their appreciation to researchers supporting project Number (RSPD2024R885) at King Saud University Riyadh Saudi Arabia for supporting this research.
References
- 1. Gibbons Johnson RM, Dong H. Functional expression of programmed death-ligand 1 (B7-H1) by immune cells and tumor cells. Frontiers in immunology. 2017;8:961. pmid:28848559
- 2. Núñez Abad M, Calabuig-Fariñas S, Lobo de Mena M, Torres-Martínez S, Garcia Gonzalez C, García García JÁ, et al. Programmed death-ligand 1 (PD-L1) as immunotherapy biomarker in breast Cancer. Cancers. 2022;14(2):307. pmid:35053471
- 3. Guan J, Lim KS, Mekhail T, Chang C-C. Programmed death ligand-1 (PD-L1) expression in the programmed death receptor-1 (PD-1)/PD-L1 blockade: a key player against various cancers. Archives of Pathology and Laboratory Medicine. 2017;141(6):851–61. pmid:28418281
- 4. Saresella M, Rainone V, M Al-Daghri N, Clerici M, Trabattoni D. The PD-1/PD-L1 pathway in human pathology. Current molecular medicine. 2012;12(3):259–67. pmid:22300137
- 5. Wang LL, Li ZH, Hu XH, Muyayalo KP, Zhang YH, Liao AH. The roles of the PD‐1/PD‐L1 pathway at immunologically privileged sites. American journal of reproductive immunology. 2017;78(2):e12710. pmid:28585775
- 6. Zitvogel L, Kroemer G. Targeting PD-1/PD-L1 interactions for cancer immunotherapy. Taylor & Francis; 2012. p. 1223–5. pmid:23243584
- 7. Salmaninejad A, Valilou SF, Shabgah AG, Aslani S, Alimardani M, Pasdar A, et al. PD‐1/PD‐L1 pathway: Basic biology and role in cancer immunotherapy. Journal of cellular physiology. 2019;234(10):16824–37. pmid:30784085
- 8. Wu B, Sun X, Gupta HB, Yuan B, Li J, Ge F, et al. Adipose PD-L1 modulates PD-1/PD-L1 checkpoint blockade immunotherapy efficacy in breast cancer. Oncoimmunology. 2018;7(11):e1500107. pmid:30393583
- 9. Tang F, Zheng P. Tumor cells versus host immune cells: whose PD-L1 contributes to PD-1/PD-L1 blockade mediated cancer immunotherapy? Cell & bioscience. 2018;8:1–8. pmid:29744030
- 10. Alsaab HO, Sau S, Alzhrani R, Tatiparti K, Bhise K, Kashaw SK, et al. PD-1 and PD-L1 checkpoint signaling inhibition for cancer immunotherapy: mechanism, combinations, and clinical outcome. Frontiers in pharmacology. 2017;8:561. pmid:28878676
- 11. Ohaegbulam KC, Assal A, Lazar-Molnar E, Yao Y, Zang X. Human cancer immunotherapy with antibodies to the PD-1 and PD-L1 pathway. Trends in molecular medicine. 2015;21(1):24–33. pmid:25440090
- 12. Massard C, Gordon MS, Sharma S, Rafii S, Wainberg ZA, Luke J, et al. Safety and efficacy of durvalumab (MEDI4736), an anti—programmed cell death ligand-1 immune checkpoint inhibitor, in patients with advanced urothelial bladder cancer. Journal of Clinical Oncology. 2016;34(26):3119. pmid:27269937
- 13. Kim ES. Avelumab: first global approval. Drugs. 2017;77(8):929–37. pmid:28456944
- 14. Durvalumab N. Trends in clinical development for PD-1/PD-L1 inhibitors. naTure reviews| Drug Discovery. 2020;19:163. pmid:32127660
- 15. Hamanishi J, Mandai M, Ikeda T, Minami M, Kawaguchi A, Murayama T, et al. Safety and antitumor activity of anti—PD-1 antibody, nivolumab, in patients with platinum-resistant ovarian cancer. Journal of Clinical Oncology. 2015;33(34):4015–22. pmid:26351349
- 16. Reilly RM, Sandhu J, Alvarez-Diez TM, Gallinger S, Kirsh J, Stern H. Problems of delivery of monoclonal antibodies: pharmaceutical and pharmacokinetic solutions. Clinical pharmacokinetics. 1995;28:126–42.
- 17. Zhan M-M, Hu X-Q, Liu X-X, Ruan B-F, Xu J, Liao C. From monoclonal antibodies to small molecules: the development of inhibitors targeting the PD-1/PD-L1 pathway. Drug Discovery Today. 2016;21(6):1027–36. pmid:27094104
- 18. Liu C, Seeram NP, Ma H. Small molecule inhibitors against PD-1/PD-L1 immune checkpoints and current methodologies for their development: A review. Cancer cell international. 2021;21(1):1–17.
- 19.
Pacheco AB, Hpc LJBRLSU. Introduction to AutoDock and AutoDock tools. 2012.
- 20. Dwivedi VD, Singh A, El-Kafraway SA, Alandijany TA, Faizo AA, Bajrai LH, et al. Mechanistic insights into the Japanese encephalitis virus RNA dependent RNA polymerase protein inhibition by bioflavonoids from Azadirachta indica. 2021;11(1):1–13.
- 21. Lagorce D, Maupetit J, Baell J, Sperandio O, Tufféry P, Miteva MA, et al. The FAF-Drugs2 server: a multistep engine to prepare electronic chemical compound collections. 2011;27(14):2018–20.
- 22. Trott O, Olson A JJJocc. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. 2010;31(2):455–61. pmid:19499576
- 23. Madhavi Sastry G, Adzhigirey M, Day T, Annabhimoju R, Sherman W JJoc-amd. Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments. 2013;27(3):221–34. pmid:23579614
- 24. Sadeer NB, Llorent-Martínez EJ, Bene K, Mahomoodally MF, Mollica A, Sinan KI, et al. Chemical profiling, antioxidant, enzyme inhibitory and molecular modelling studies on the leaves and stem bark extracts of three African medicinal plants. 2019;174:19–33.
- 25. Shivakumar D, Harder E, Damm W, Friesner RA, Sherman W JJoct, computation. Improving the prediction of absolute solvation free energies using the next generation OPLS force field. 2012;8(8):2553–8. pmid:26592101
- 26. Matsuoka M, Kumar A, Muddassar M, Matsuyama A, Yoshida M, Zhang KY JJoCI, et al. Discovery of fungal denitrification inhibitors by targeting copper nitrite reductase from Fusarium oxysporum. 2017;57(2):203–13. pmid:28117584
- 27.
Bowers KJ, Chow E, Xu H, Dror RO, Eastwood MP, Gregersen BA, et al., editors. Scalable algorithms for molecular dynamics simulations on commodity clusters. Proceedings of the 2006 ACM/IEEE Conference on Supercomputing; 2006.
- 28. Price DJ, Brooks CL III JTJocp. A modified TIP3P water potential for simulation with Ewald summation. 2004;121(20):10096–103. pmid:15549884
- 29. Kumar S JJoBWorld Ts. Comparative Modeling, Virtual Screening and Docking of CCR5 with Diverse Libraries: An Initiation towards Rational Drug Design to Identify a New Chemical Entity to Treat HIV. 2019;8(7):1–4.
- 30.
Visualizer DJAsi. Discovery Studio Visualizer. 2. 2005.
- 31. Thillainayagam M, Malathi K, Ramaiah S JJoBS, Dynamics. In-Silico molecular docking and simulation studies on novel chalcone and flavone hybrid derivatives with 1, 2, 3-triazole linkage as vital inhibitors of Plasmodium falciparum dihydroorotate dehydrogenase. 2018;36(15):3993–4009. pmid:29132266
- 32. Sargsyan K, Grauffel C, Lim C JJoct, computation. How molecular size impacts RMSD applications in molecular dynamics simulations. 2017;13(4):1518–24. pmid:28267328
- 33. Martínez L JPo. Automatic identification of mobile and rigid substructures in molecular dynamics simulations and fractional structural fluctuation analysis. 2015;10(3):e0119264.
- 34. Zhang B, Song Y, Fu Y, Zhu B, Wang B, Wang J. Current status of the clinical use of PD-1/PD-L1 inhibitors: a questionnaire survey of oncologists in China. BMC cancer. 2020;20:1–7. pmid:32005140
- 35. Ai L, Chen J, Yan H, He Q, Luo P, Xu Z, et al. Research status and outlook of PD-1/PD-L1 inhibitors for cancer therapy. Drug design, development and therapy. 2020:3625–49. pmid:32982171
- 36. Yang J, Hu L. Immunomodulators targeting the PD‐1/PD‐L1 protein‐protein interaction: from antibodies to small molecules. Medicinal research reviews. 2019;39(1):265–301. pmid:30215856
- 37. Ganesan A, Ahmed M, Okoye I, Arutyunova E, Babu D, Turnbull WL, et al. Comprehensive in vitro characterization of PD-L1 small molecule inhibitors. Scientific reports. 2019;9(1):12392. pmid:31455818
- 38. Shaabani S, Huizinga HP, Butera R, Kouchi A, Guzik K, Magiera-Mularz K, et al. A patent review on PD-1/PD-L1 antagonists: small molecules, peptides, and macrocycles (2015–2018). Expert opinion on therapeutic patents. 2018;28(9):665–78. pmid:30107136
- 39. Sasikumar PG, Ramachandra M. Small-molecule immune checkpoint inhibitors targeting PD-1/PD-L1 and other emerging checkpoint pathways. BioDrugs. 2018;32(5):481–97. pmid:30168070
- 40. Scott DE, Bayly AR, Abell C, Skidmore J. Small molecules, big targets: drug discovery faces the protein—protein interaction challenge. Nature Reviews Drug Discovery. 2016;15(8):533–50. pmid:27050677
- 41. Jang GR, Harris RZ, Lau DT. Pharmacokinetics and its role in small molecule drug discovery research. Medicinal research reviews. 2001;21(5):382–96. pmid:11579439
- 42. Rao VS, Srinivas K. Modern drug discovery process: An in silico approach. Journal of bioinformatics and sequence analysis. 2011;2(5):89–94.
- 43. Shaker B, Ahmad S, Lee J, Jung C, Na D. In silico methods and tools for drug discovery. Computers in biology and medicine. 2021;137:104851. pmid:34520990
- 44.
Parmar G, Chudasama JM, Shah A, Patel A. In Silico Pharmacology and Drug Repurposing Approaches. CADD and Informatics in Drug Discovery: Springer; 2023. p. 253–81.
- 45. Shah A, Choudhary A, Jain M, Perumal S, Patel V, Parmar G, et al. Discovery of novel anticancer flavonoids as potential HDAC2 inhibitors: virtual screening approach based on molecular docking, DFT and molecular dynamics simulations studies. 3 Biotech. 2024;14(3):1–20.
- 46. Shah A, Parmar G, Shah U, Perumal S. Virtual Screening, Molecular Docking Studies and DFT Calculations of Novel Anticancer Flavonoids as Potential VEGFR-2 Inhibitors. Chemistry Africa. 2023;6(4):1847–61.
- 47. Zhou H, Wang C, Ye J, Chen H, Tao R. Design, virtual screening, molecular docking and molecular dynamics studies of novel urushiol derivatives as potential HDAC2 selective inhibitors. Gene. 2017;637:63–71. pmid:28939339
- 48. Jakhar R, Dangi M, Khichi A, Chhillar AK. Relevance of molecular docking studies in drug designing. Current Bioinformatics. 2020;15(4):270–8.
- 49. Jia C-Y, Li J-Y, Hao G-F, Yang G-F. A drug-likeness toolbox facilitates ADMET study in drug discovery. Drug discovery today. 2020;25(1):248–58. pmid:31705979
- 50. Shah AP, Parmar BM, Ghodawala MA, Seth AJJoIHS. In silico drug discovery of novel small lead compounds targeting nipah virus attachment glycoprotein. 2018;6(2):60–3.
- 51. Parmar G, Shah A, Shah S, Seth AK JBRAC. Identification of bioactive phytoconstituents from the plant Euphorbia hirta as potential inhibitor of SARS-CoV-2: an in-silico approach. 2022;12:1385–96.
- 52. Salo-Ahen OM, Alanko I, Bhadane R, Bonvin AM, Honorato RV, Hossain S, et al. Molecular dynamics simulations in drug discovery and pharmaceutical development. Processes. 2020;9(1):71.
- 53.
Rudrapal M, Egbuna C. Computer aided drug design (CADD): From ligand-based methods to structure-based approaches: Elsevier; 2022.
- 54. Zhao L, Ciallella HL, Aleksunes LM, Zhu H. Advancing computer-aided drug discovery (CADD) by big data and data-driven machine learning modeling. Drug discovery today. 2020;25(9):1624–38. pmid:32663517