Computational and experimental design of l-amino acid-based alternatives to ketorolac Authors: Morteza Hosseini, Alireza Fattahi, Atena Pakzadiyan Published: 2025 ABSTRACT l-amino acids play a vital role in various biological processes, including hormone production, adipose tissue growth and repair, and protein synthesis. Consequently, the design of novel l-amino acid-based drugs with fewer side effects has recently garnered significant attention. Ketorolac, while effective for pain reduction, presents several adverse side effects, motivating the search for alternatives. An ideal alternative should structurally resemble ketorolac but offer improved chemical and thermal stability, as well as enhanced hydrogen bonding capability. In this project, we designed an l-amino acid-based structure as a potential alternative to ketorolac. Using computer-aided drug discovery methods, we generated a detailed, atomic-level description of the drug ligand and evaluated its selectivity properties, ADMET prediction, and Pre-Tox analysis. Molecular docking calculations were then performed to identify the most favorable active site for the newly designed ligand, mimicking ketorolac’s binding pose within the protein. The resulting AVH structure exhibited a close resemblance to ketorolac. After these initial assessments, we performed molecular dynamics simulations. The RMSD and RMSF plots indicate the stability of the dynamic structure of the substituted ligand. Analysis of the rotation and hydrogen bond plots revealed strong interactions between the protein and the ligand. Furthermore, analysis of the protein secondary structure and residue-wise secondary structure using the DSSP algorithm confirmed the suitability of the proposed structure and, in some cases, showed improved stability compared to its interaction with ketorolac. In the DCCM analysis, a stronger correlation pattern was observed in AVH, indicating greater stability in this pattern. Principal component analysis based on a free energy perspective revealed closely spaced data points representing low-energy and relatively stable states that are energetically favorable for the system. The protein binding cavity in the electrostatic potential map shows negative and neutral regions, likely indicating greater stabilization of the ligand in the active site. Finally, the MM-PBSA results suggest that although the binding energy of AVH is slightly higher than that of KTR, it remains a negative value, which is thermodynamically favorable, and the binding affinity of AVH to the protein remains strong. Based on these results, we selected and synthesized the best structure compared to ketorolac. We predict that this new structure, designed using a natural product and showing similar chemical effects, will be a potential alternative to ketorolac.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-26171-1. Introduction l-amino acids and their derivatives are natural products and valuable candidates for drug design. As the building blocks of peptides and proteins, key components are involved in vital bodily functions. l-amino acids possess a wide range of specific physiological activities and play a crucial role in protein synthesis. Beyond these functions, l-amino acids are extensively used in foods, pharmaceuticals, and supplements; for example, l-phenylalanine serves as the raw material for the sweetener aspartame. They are also widely utilized in the medical and healthcare industries. Depending on their properties, l-amino acids exhibit a diverse range of pharmacological activities. One study reported that conjugating a molecule with an amino acid significantly enhanced its pharmacological efficacy, reduced cytotoxicity, and improved solubility. l-amino acids can improve the efficacy, stability, and bioavailability of pharmaceutical agents due to their distinct physical and chemical properties. Additionally, they can significantly enhance the solubility of poorly soluble drugs, which is crucial for optimal absorption in the body, as increased solubility typically leads to improved therapeutic effects. By incorporating l-amino acids into pharmaceutical formulations, the integrity of the drug composition can be maintained, and their absorption into the systemic circulation can be optimized1–12. Therefore, the research and development of novel pharmaceuticals derived from natural substances have garnered significant interest in recent years. Recently, structures utilizing l-amino acids and sugars are being increasingly engineered as enzyme inhibitors and therapeutic agents. These compounds have piqued the interest of research teams and pharmaceutical companies due to their cost-effectiveness and accessibility. The creation of new structures based on l-amino acids presents a promising avenue for synthesizing drugs with varied properties tailored to specific biological functions. Given their derivation from natural compounds, these structures are expected to exhibit fewer side effects. This potential, along with the previously mentioned advantages, has recently captured the attention of numerous research groups, including our own. Kalhor et al. have utilized quantum mechanics (QM) analyses and molecular dynamics (MD) simulations to develop anticancer agents targeting polymerase η, a protein essential for DNA replication, translation, and repair. The integration of l-amino acids with sugars or small molecules has been demonstrated to influence various factors, including metabolic stability, solubility, membrane permeability, biodistribution, and ligand-target interactions13–25.Moreover, l-amino acid nucleobases can also be utilized to design and synthesize peptides that play a crucial role in the binding of DNA and RNA. Specific interactions, such as hydrogen bonding between the nucleobase in the peptide and the base in the RNA, can enhance the binding affinity and specificity for DNA and RNA26,27. l-amino acids are often used as carriers in prodrug design to enhance the bioavailability and targeting of therapeutic agents that are poorly absorbed. Attaching l-amino acid moieties to drugs can improve their transport across biological membranes and reduce their toxicity. This strategy leverages endogenous nutrient transporters that recognize l-amino acid derivatives, thus facilitating targeted delivery to specific tissues or organs28,29. l-amino esters are crucial intermediates in organic synthesis, particularly in the preparation of peptides. Various methods, each with distinct mechanisms and conditions, can be employed to synthesize these compounds. One common approach involves the direct esterification of l-amino acids with alcohols in the presence of acid catalysts. For example, heating an l-amino acid with an alcohol in the presence of sulfuric acid or p-TSA catalysts can efficiently produce the corresponding l-amino esters30–32. Amino esters can also be prepared from l-amino acids in the presence of reagents such as thionyl chloride and oxalyl chloride, with high efficiency and yield33,34. l-amino esters can be used as biomimetic reducing agents to reduce carbon–carbon double bonds in various functional groups, including oxindoles, under mild conditions with good yields35. l-amino esters play a significant role in drug development, particularly as prodrugs that enhance the pharmacological properties of various compounds. Several drugs and natural compounds are known to utilize l-amino esters, such as valacyclovir, an antiviral drug linked to valine via a glycol linker, which enhances its activity against herpes viruses and the natural lignan arctigenin, isolated from Forsythia koreana, which has antitumor activity against a variety of cancers, including pancreatic, gastric, colon, and liver cancer36,37. l-amino esters have also been recently investigated as ionic liquid vehicles for nonsteroidal anti-inflammatory drugs in potential local drug delivery systems with antimicrobial activity38. Nonsteroidal anti-inflammatory drugs (NSAIDs) are a class of widely used medications that significantly reduce inflammatory factors in the body. They achieve these effects primarily by inhibiting cyclooxygenase (COX) enzymes, which are involved in the synthesis of prostaglandins. Ketorolac (KTR) is a nonsteroidal anti-inflammatory drug (NSAID) with potent analgesic activity. The analgesic effect of this drug has been extensively evaluated in inpatients, outpatients, and patients with other acute pain conditions. Although ketorolac has been reported to inhibit cyclooxygenase-2 (COX-2) activity and affect arachidonic acid metabolism, it is mostly known as a nonselective COX inhibitor of a heteroaryl acid that has a chiral center and exists as a racemic mixture of the (−)S and (+)R enantiomers. The anti-inflammatory and analgesic activity of this drug is primarily associated with the (−)S configuration, which also exhibits ulcerogenic activity. The analgesic effect of ketorolac depends on the concentration of the (−)S and (+)R enantiomers in the racemic mixture. Most clinically important adverse events affect the gastrointestinal tract or renal function. The incidence of serious or fatal adverse events reported with KTR has recently decreased. Results from a study involving a small number of patients showed that the overall risk of gastrointestinal or surgical site bleeding associated with intravenous KTR treatment was only slightly higher than for other sensitivities to this drug. Like other NSAIDs, KTR has been implicated in allergic or hypersensitivity reactions. In patients experiencing moderate to severe pain, it serves as a useful alternative or adjunct to opioids39–41. In previous studies reported by Annalisa Curcio et al. and Roberto Russo et al42,43., it has been observed that the esterification reaction of d-galactose with ketorolac (Ketogal) significantly reduced the major side effect of the drug, which is gastrointestinal pain (gastric ulcers), in addition to maintaining its anti-inflammatory properties. The interesting point here is the conversion of the acidic functional group to ester by d-galactose, which led to this promising result. As mentioned above, hypersensitivity, gastrointestinal problems, acute renal failure, and subsequent nausea are the most important side effects of ketorolac (KTR). Therefore, the primary goal of this research project is to design a compound that produces a comparable anti-inflammatory effect and exhibits fewer side effects than KTR. In this study, we employed molecular docking, molecular dynamics simulation, ADMET prediction, Pre-Tox analysis and MM-PBSA calculations to design and select an Fig. 1The structure of (a) Ketorolac (KTR), (b) prop‐2‐en‐1‐yl (2S)‐2‐amino‐3 methyl butanoate hydrochloride (AVH). The structure of (a) Ketorolac (KTR), (b) prop‐2‐en‐1‐yl (2S)‐2‐amino‐3 methyl butanoate hydrochloride (AVH). Method and theoretical calculations The structural model First, the possible structures of allyl alcohol and l-amino acids were obtained using MarvinSketch44,45. Since there are approximately 500 l-amino acids in nature, only 20 l-amino acids constitute proteins found in the human body. Considering the number of l-amino acids and allyl alcohol (as the alcohol agent), a total of 20 possible structures were estimated. The proposed structures are shown in Table 1. After examining the number of possible structures using MarvinSketch, considering that each structure can have multiple molecular conformers, we used Spartan46,47 (molecular mechanics with MM force field) software to obtain the best conformer (the most stable conformer). Subsequently, initial conformation searches were performed using the MMFF force field (molecular mechanics) with Spartan software to obtain the best conformer for energy optimization calculations. Among the three force fields defined in the software, MMFF, MMFFaq, and SYBYL, we used MMFF in this study. Compared to SYBYL, MMFF has a wider parameterization for different molecules, higher computational speed and efficiency, as well as greater compatibility and more reliable results. Additionally, since some of the proposed compounds exhibit different solubilities in aqueous systems, MMFFaq, which performs calculations in aqueous environments, was not investigated. As mentioned, calculations were performed with MMFF to conduct a more accurate and comprehensive investigation. The physical properties, including volume, surface area, and polarizability, of the drug and the designed structures were compared. These results (listed in Table 1) were obtained by single-point energy calculations on the optimized structures at the REDF2/6-31G(d) level (in vacuum) using Spartan software. As shown in Table 1, sulfur-based and essential L-amino acids generally exhibit less similarity, whereas neutral and acidic L-amino acids display more remarkable similarity to the parent drug. On the other hand, acidic L-amino acids such as glutamic acid and aspartic acid, despite having similar physical properties (area, volume) and affinity for the drug, may produce mono, diesters, or other by-products due to the presence of two acidic groups in the synthesis process, which are not of industrial value. Additionally, the diester form, having one more non-polar chain, may not possess the desired solubility in water compared to neutral L-amino acids. Therefore, the AVH structure (entry 20 in Table 1) was considered the most suitable candidate to replace KTR. Molecular docking, Molecular dynamics simulations, ADMET predictions, pre-Tox analysis, and MM-PBSA calculations were then performed for AVH and KTR to investigate their properties further. The results indicated that these two structures are very similar to each other. Finally, the structure was included in the experimental section.Table 1Information on the physical properties of the drug ketorolac and proposed designed structures based on L-type amino acids.EntryName of structureStructureAffinty (kcal/mol)Area (Å2)Volume (Å3)Polarizability1Prop‐2‐en‐1‐yl (2S)‐2‐aminopropanoate hydrochloride− 4.8208.61167.8154.082prop‐2‐en‐1‐yl(2S) ‐2amino‐5-Carbamimidamidopentanoate dihydrochloride− 6.1322.26274.0762.393prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐Carbamoylpropanoate hydrochloride− 5.3240.73198.2856.4341,4‐bis(prop‐2‐en‐1‐yl) (2S) ‐2‐Aminobutanedioate hydrochloride− 6.1274.09230.0959.545prop‐2‐en‐1‐yl (2S) ‐2‐amino‐3‐Sulfanylpropanoate hydrochloride− 4.6223.83180.01555.0861,5‐bis(prop‐2‐en‐1‐yl) (2S) ‐2‐Aminopentanedioate hydrochloride− 6.2312.21266.1462.027prop‐2‐en‐1‐yl (2S)‐2‐amino‐4‐Carbamoylbutanoate hydrochloride− 5.9216.06258.0258.298Prop‐2‐en‐1‐yl 2‐aminoacetate hydrochloride− 4.2190.01149.5552.599prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐(1H‐imidazol‐4‐yl) propanoate hydrochloride− 6.0265.38225.5658.7710Prop‐2‐en‐1‐yl (2S,3S)‐2‐amino‐3‐methylpentanoate hydrochloride− 5.7264.91251.6160.8711prop‐2‐en‐1‐yl (2S)‐2‐amino‐4‐methylpentanoate hydrochloride− 6.0266.55222.6958.5312Prop‐2‐en‐1‐yl (2S)‐2,6‐diaminohexanoate dihydrochloride− 4.4301.76255.7460.8813Prop‐2‐en‐1‐yl (2S)‐2‐amino‐4‐(methylsulfanyl)butanoate hydrochloride− 4.9265.47223.1858.5614Prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐phenylpropanoate hydrochloride− 6.8287.41251.6160.8715Prop‐2‐en‐1‐yl (2S)‐pyrrolidine‐2‐carboxylate hydrochloride− 5.9231.67194.4956.2416Prop‐2‐en‐1‐yl (2S) ‐2‐amino‐3‐hydroxypropanoate hydrochloride− 4.5217.05175.3254.6917Prop‐2‐en‐1‐yl (2S,3S)‐2‐amino‐3‐hydroxybutanoate hydrochloride− 5.1235.28193.5256.1718Prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐(1H‐indol‐3‐yl) propanoate hydrochloride6.9319.21284.2663.5019Prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐(4‐hydroxyphenyl) propanoate hydrochloride− 6.7296.34258.6961.4320Prop‐2‐en‐1‐yl (2S) ‐2‐amino‐3‐methylbutanoate hydrochloride (AVH)− 6.6246.18204.3457.0121(1R)‐5‐benzoyl‐2,3‐dihydro‐1H‐pyrrolizine‐1‐carboxylic acid (KTR)− 8.7281.55261.9661.65 Table 1Information on the physical properties of the drug ketorolac and proposed designed structures based on L-type amino acids.EntryName of structureStructureAffinty (kcal/mol)Area (Å2)Volume (Å3)Polarizability1Prop‐2‐en‐1‐yl (2S)‐2‐aminopropanoate hydrochloride− 4.8208.61167.8154.082prop‐2‐en‐1‐yl(2S) ‐2amino‐5-Carbamimidamidopentanoate dihydrochloride− 6.1322.26274.0762.393prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐Carbamoylpropanoate hydrochloride− 5.3240.73198.2856.4341,4‐bis(prop‐2‐en‐1‐yl) (2S) ‐2‐Aminobutanedioate hydrochloride− 6.1274.09230.0959.545prop‐2‐en‐1‐yl (2S) ‐2‐amino‐3‐Sulfanylpropanoate hydrochloride− 4.6223.83180.01555.0861,5‐bis(prop‐2‐en‐1‐yl) (2S) ‐2‐Aminopentanedioate hydrochloride− 6.2312.21266.1462.027prop‐2‐en‐1‐yl (2S)‐2‐amino‐4‐Carbamoylbutanoate hydrochloride− 5.9216.06258.0258.298Prop‐2‐en‐1‐yl 2‐aminoacetate hydrochloride− 4.2190.01149.5552.599prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐(1H‐imidazol‐4‐yl) propanoate hydrochloride− 6.0265.38225.5658.7710Prop‐2‐en‐1‐yl (2S,3S)‐2‐amino‐3‐methylpentanoate hydrochloride− 5.7264.91251.6160.8711prop‐2‐en‐1‐yl (2S)‐2‐amino‐4‐methylpentanoate hydrochloride− 6.0266.55222.6958.5312Prop‐2‐en‐1‐yl (2S)‐2,6‐diaminohexanoate dihydrochloride− 4.4301.76255.7460.8813Prop‐2‐en‐1‐yl (2S)‐2‐amino‐4‐(methylsulfanyl)butanoate hydrochloride− 4.9265.47223.1858.5614Prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐phenylpropanoate hydrochloride− 6.8287.41251.6160.8715Prop‐2‐en‐1‐yl (2S)‐pyrrolidine‐2‐carboxylate hydrochloride− 5.9231.67194.4956.2416Prop‐2‐en‐1‐yl (2S) ‐2‐amino‐3‐hydroxypropanoate hydrochloride− 4.5217.05175.3254.6917Prop‐2‐en‐1‐yl (2S,3S)‐2‐amino‐3‐hydroxybutanoate hydrochloride− 5.1235.28193.5256.1718Prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐(1H‐indol‐3‐yl) propanoate hydrochloride6.9319.21284.2663.5019Prop‐2‐en‐1‐yl (2S)‐2‐amino‐3‐(4‐hydroxyphenyl) propanoate hydrochloride− 6.7296.34258.6961.4320Prop‐2‐en‐1‐yl (2S) ‐2‐amino‐3‐methylbutanoate hydrochloride (AVH)− 6.6246.18204.3457.0121(1R)‐5‐benzoyl‐2,3‐dihydro‐1H‐pyrrolizine‐1‐carboxylic acid (KTR)− 8.7281.55261.9661.65 Information on the physical properties of the drug ketorolac and proposed designed structures based on L-type amino acids. Molecular docking Molecular docking studies were performed using AutoDock Vina48 to assess the interaction capabilities of the compounds as ligands and to generate starting coordinate setups for Molecular Dynamics (MD) simulations. The protein structure was prepared using AutoDock Tools49 by incorporating polar hydrogens and assigning Kollman charges as fractional charges before being stored in the pdbqt format. Similarly, the compound structures, with added polar hydrogens and computed partial charges, were also saved as pdbqt files. The binding affinity between the known compound, KTR, and Protein PDBID = 2MJI (http://www.pdb.org) was considered a reference, and the binding affinity of the proposed structures with the Protein was compared with that of KTR. To calculate the binding affinity between the Protein and each drug molecule, a molecular docking study was performed using AutoDock Vina4.2. A grid box with dimensions 6 × 11 × 8 Å, centered at (0.789, − 0.324, − 5.062), was used for docking. The Broyden–Fletcher–Goldfarb–Shanno (BFGS) algorithm, with an exhaustiveness setting of 200, was employed for local optimization during the docking run. The best ligand-binding orientations were then selected based on energy scores and spatial arrangement to achieve preferred interactions. ADMET prediction and Pre-Tox analysis In the initial phases of drug design and discovery, leveraging ADMET prediction to investigate physicochemical characteristics, toxicological profiles, and interactions with biological systems and cellular metabolism is critical for guiding subsequent drug development and regulatory approval processes. Accordingly, this study employed in silico computational approaches using DataWarrior50 software to evaluate and compare the physicochemical, pharmacokinetic, and pharmacodynamic parameters collectively referred to as ADMET of both compounds. The results show that the proposed compound complies with the five Lipinski rules, indicating its potential as an excellent alternative candidate. To ensure compatibility with the software, the molecular structures were converted from PDB to SDF format using the software Obabel v. 2.3 (http://openbabel.org)51. For advanced appraisal,we used Pre-Tox analysis (https://tox.charite.de/protox3/)52, an advanced computational platform that complements ADMET predictions by providing a comprehensive and accurate assessment of toxicity profiles. This tool integrates fragment-based analysis, molecular similarity measures, and machine learning algorithms to predict diverse toxicological endpoints with high accuracy. Molecular dynamics (MD) simulations Molecular dynamics (MD) simulations were performed using the GROMACS 2019 and 202353 software suites. GROMACS 2019 was used during the initial setup of the system and for specific analyses such as DSSP (Definition of Secondary Structure of Proteins). We observed that the gmx dssp module in the 2019 version provided more consistent and interpretable results. In contrast, GROMACS 2023 introduced changes to the DSSP interface that limited this capability. Therefore, GROMACS 2023 was used for the final molecular dynamics simulations, leveraging its performance and GPU acceleration capabilities. These simulations were designed to investigate the dynamic behavior and interaction profiles of docking complexes, with a specific focus on the binding of a ligand to a target protein. The visual molecular dynamics VMD 1.9.1 (https://www.ks.uiuc.edu/) program was used for molecular visualization54. Ligand parameterization was achieved using the General Amber Force Field55 employing the acpype tool56 from the bio2byte server (https://www.bio2byte.be/acpype/)57. The Amber99SB force field was employed to generate topology files and protein coordinates, which were processed using the pdb2gmx module within GROMACS. The protein–ligand complex was solvated in a cubic simulation box (6 × 6 × 4 nm3) with TIP3P water molecules. We chose the TIP3P three-point water model due to its computational efficiency, compatibility with the chosen force field, and explicit representation of solvent molecules. Although more accurate models, such as TIP4P or TIP5P, offer improved accuracy in reproducing certain thermodynamic properties of water, they also increase computational cost and complexity. System charge neutralization was accomplished by adding counterions. Initial energy minimization was performed using the steepest descent algorithm to alleviate steric clashes and reduce the system’s potential energy. Following energy minimization, a two-stage equilibration process was implemented. Positional restraints were applied to the protein and ligand during a 2 ns NVT ensemble simulation, followed by a 2 ns NPT ensemble simulation. System equilibrium was assessed by monitoring thermodynamic parameters, including pressure, density, potential energy, and temperature. Temperature and pressure were maintained at 310 K and 1 bar, respectively, using the Berendsen thermostat and Parrinello-Rahman barostat58. The LINCS algorithm was used to constrain bond lengths involving hydrogen atoms59. Long-range electrostatic interactions were calculated using the Particle Mesh Ewald60 (PME) method. Production MD simulations were run for 200 ns in the extended simulation cases, during which the atoms in the system were allowed to evolve according to Newton’s equations of motion. It is worth noting that we initially performed several 100 ns simulations to optimize the molecular systems before the initial investigations. To ensure the statistical accuracy of the results, we performed three independent 200ns iterative molecular dynamics simulations for each system, strictly adhering to the same simulation protocols and parameters. The results from these iterative simulations consistently show that AVH exhibits a completely distinct dynamic stability profile compared to KTR. MM-PBSA free energy analysis The free energy (ΔG) was calculated using the Molecular Mechanics/Poisson-Boltzmann Surface Area (MM-PBSA) method61. This analysis is a reliable computational method for estimating the free energy between molecules by combining molecular mechanics and continuum solvation models based on molecular dynamics simulations. Although the MM-GBSA method is computationally faster, it relies on approximations that can reduce accuracy. The MM-PBSA method provides a more accurate and reliable estimate of the polar solvation energy by solving the PB equation, which was very important considering the polar interactions present in our system. It has higher sensitivity and accuracy for the target systems in molecular dynamics simulations. Therefore, considering the importance of high sensitivity and accuracy, as well as polar interactions in the system under study, we decided to use MM-PBSA for calculating the binding energy. All computations were performed under the standard configuration settings supplied by the developers of the gmx_mmpbsa packages at 310 K over the final 200 ns of equilibrated trajectories. Results and discussion Molecular docking Molecular docking experiments were conducted utilizing the crystal structure of the protein complexed with KTR (PDB ID: 2MJI)62. The structure of fatty acid binding protein is shown in Fig. 2.Fig. 2The protein structure is visually depicted with beta-sheets in red, alpha-helices in blue, and loops or turns in grey. Side chains are represented as sticks or ball-and-stick models. The protein structure is visually depicted with beta-sheets in red, alpha-helices in blue, and loops or turns in grey. Side chains are represented as sticks or ball-and-stick models. Initially, the best conformer of both structures was optimized using the M062x/6-31++ G(d,p) level of theory63 (Table 2). The preparation of both the Protein and ligand structures were carried out according to the methodology detailed in the Methods section. AutoDock Vina was employed to estimate the binding strength of the ligands within the receptor’s active site. This step also produced the starting coordinate files necessary for subsequent molecular dynamics simulations. To validate the docking protocol and enable a comparison between AVH and KTR, a redocking procedure was performed within the protein’s binding pocket.Table 2Optimized geometries, obtained at M062x/6-31++ G(d,p) level for the ground-state configurations of KTR and AVH structures, detailing the positions of all atoms.KTRAVHATOM Table 2Optimized geometries, obtained at M062x/6-31++ G(d,p) level for the ground-state configurations of KTR and AVH structures, detailing the positions of all atoms.KTRAVHATOM Optimized geometries, obtained at M062x/6-31++ G(d,p) level for the ground-state configurations of KTR and AVH structures, detailing the positions of all atoms. Figures 3 and S1 illustrate a comparative analysis of the l-amino acid interactions between the protein and KTR versus AVH. The AVH complex exhibited interactions with six residues within the contact region (Tyr70, Phe17, Phe93, Met21, Leu102, Tyr117) characterized by one hydrogen bond.Fig. 3(a) The most favorable binding interaction of KTR within the receptor’s binding pocket, as determined from 200 independent docking simulations using AutoDock. (b) The optimal binding pose of AVH. (a) The most favorable binding interaction of KTR within the receptor’s binding pocket, as determined from 200 independent docking simulations using AutoDock. (b) The optimal binding pose of AVH. Inhibition constant (Ki)64 is a fundamental parameter for describing inhibitory effects that is very important in early and late stages of drug development because it determines the affinity of an inhibitor (potential drug) for its target enzyme or receptor. Ki is also used to assess the effectiveness of a drug in inhibiting an enzyme that is involved in metabolism or disease progression. It is calculated as Ki = exp(ΔG/(R × T)) where ΔG is the free energy of binding, R is the gas constant (1.987 cal K−1 mol−1) and T is the temperature (298.15 K). In addition, ligand efficiency65 is another parameter of Binding Efficiency Metrics. LE is a normalized measure of binding energy that takes into account the size of the ligand and is usually expressed per heavy atom (non-hydrogen atom). This value is calculated as follows: LE = − ΔG/N where ΔG is the free energy of binding and N is the number of heavy atom (non-hydrogen atom). Table 3 shows the binding energies for both compounds studied, including the KTR compound which has the strongest binding at − 8.7 and the AVH compound at − 6.6 kcal/mol as calculated by AutoDock Vina. Binding efficiency metrics, including inhibition constant (Ki) and ligand efficiency (LE) values normalized by the number of heavy atoms, are also reported for both compounds.Table 3The binding energy of the active site docking of the protein calculated by AutoDock Vina as well as binding efficiency metrics, including inhibition constant (Ki) and ligand efficiency (LE) values normalized by the number of heavy atoms, are also reported for both KTR and AVH compounds.KTRAVHAffinity to binding (kcal/mol)− 8.7− 6.6Estimated inhibition constant (Ki)4.16 × 10–71.44 × 10–5Ligand efficiency (LE)0.460.55 Table 3The binding energy of the active site docking of the protein calculated by AutoDock Vina as well as binding efficiency metrics, including inhibition constant (Ki) and ligand efficiency (LE) values normalized by the number of heavy atoms, are also reported for both KTR and AVH compounds.KTRAVHAffinity to binding (kcal/mol)− 8.7− 6.6Estimated inhibition constant (Ki)4.16 × 10–71.44 × 10–5Ligand efficiency (LE)0.460.55 The binding energy of the active site docking of the protein calculated by AutoDock Vina as well as binding efficiency metrics, including inhibition constant (Ki) and ligand efficiency (LE) values normalized by the number of heavy atoms, are also reported for both KTR and AVH compounds. ADMET prediction and Pre-Tox analysis The DataWarrior software predicted key physicochemical properties, including pharmacokinetics and toxicity risk, for the compounds, and the results are presented in Table 4. Lipinski’s Rule of Five (Ro5) can assess whether these compounds possess suitable physicochemical properties for oral biopharmaceutical use in humans66. The Ro5 utilizes cut-off values for four descriptors: molecular weight (MW) ≤ 500 g/mol, number of hydrogen bond acceptors (HBA) ≤ 10, number of hydrogen bond donors (HBD) ≤ 5, and the partition coefficient (cLogP) ≤ 5. These guidelines aim to maximize solubility without compromising permeability. Compound AVH showed the highest compatibility with Ro5. Weber et al. found that reducing molecular flexibility, reflected by the number of rotatable bonds, and limiting the total number of hydrogen bond donors and acceptors are key factors in increasing bioavailability, regardless of molecular weight. Reducing these factors increases a compound’s ability to penetrate cell membranes because fewer hydrogen bond interactions increase lipid solubility, easing passage across the phospholipid bilayer. Fewer rotatable bonds reduce entropic costs during conformational changes, promoting better membrane permeation efficiency. Among the pharmacokinetic compounds, AVH, with a reduced molecular weight compared to KTR (from 255.27 to 158.11 g/mol), exhibits greater pharmacological similarity than KTR and is likely to have more suitable oral bioavailability as a drug in humans. Lipophilicity, described as the ratio between a compound’s solubility in octanol and water, is another key physicochemical property. The logarithmic function of this ratio (cLogP) predicts this value and is very important in pharmacology for predicting how drugs behave in the body, affecting absorption, metabolism, as well as their permeability across membranes and biological barriers67.Table 4Data Warrior calculation of physicochemical and pharmacokinetic (lipophilicity, water solubility) properties for the ADMET investigation of KTR and AVH.CompoundsMW (g/mol)No. HBANo. HBDNo. Rot*cLogPcLogSToxicity riskMolecular flexibilityKTR255.274132.36− 2.61None0.31AVH158.113150.51− 1.47None0.70*No. Rot: count of rotatable bonds. Table 4Data Warrior calculation of physicochemical and pharmacokinetic (lipophilicity, water solubility) properties for the ADMET investigation of KTR and AVH.CompoundsMW (g/mol)No. HBANo. HBDNo. Rot*cLogPcLogSToxicity riskMolecular flexibilityKTR255.274132.36− 2.61None0.31AVH158.113150.51− 1.47None0.70*No. Rot: count of rotatable bonds. Data Warrior calculation of physicochemical and pharmacokinetic (lipophilicity, water solubility) properties for the ADMET investigation of KTR and AVH. *No. Rot: count of rotatable bonds. According to Ro5, the clogP value of a chemical compound intended for oral administration should be less than 5. Compounds with higher lipophilicity may have poor water solubility. The ideal clogP value for a drug intended for oral and enteral absorption is in the range of 1.35–1.8, which is where compound AVH falls. LogS is a key parameter for characterizing a drug’s behavior in aqueous environments. It is defined as log10(S), where 'S' is the drug’s solubility in mol/L, indicating the maximum concentration achievable in water68. Also, Water solubility is a challenging parameter because it is essential to achieve the desired drug concentration in the bloodstream and create a homogeneous system for the desired drug response. Highly lipophilic compounds may be retained by adipose tissue, making elimination difficult and leading to accumulation that can affect the systemic toxicity of the substance. The toxicity risk assessment was further evaluated using Data Warrior software. This analysis sought to identify substructural elements within the examined chemical entity that are associated with known toxic effects, classifying them according to four primary toxicity categories. These problematic chemical structures possess substructural fragments that fall into these four key classes: mutagenicity, tumorigenicity, irritation, and reproductive toxicity. The underlying principle is that structural motifs frequently dictate toxicological behavior. Consequently, these substructures, predominantly observed in toxic substances but seldom in non-toxic ones, are presumed to be responsible for the observed toxicity. The toxicity risk assessment of our compounds revealed that none exhibited characteristics associated with the four major toxicity classes; hence, they were predicted to be non-toxic. Considering a broader perspective, AVH, in addition to its demonstrated biological activity, displayed the highest compatibility with Lipinski’s Rule of Five (Ro5) criteria, suggesting potentially better oral bioavailability compared to other compounds with pharmacokinetic properties. Based on the toxicity risk assessment, the investigated compound was predicted to be non-toxic and did not fall into any of the four major toxicity categories. For further validation, we performed a more comprehensive assessment of toxicity parameters using the ProTox 3.0 web server. As mentioned, this analysis is based on computational models from machine learning, fragment analysis, and molecular similarity that estimate toxicity effects using in silico models that can be a good complement to ADMET predictions. The results obtained from ProTox 3.0 for both compounds have been added to the modified backup version of the table. Therefore, the compounds KTR and AVH were plotted separately on the server itself, and each was entered into the web server. The results showed that both compounds are of the predicted toxicity class 3. The median lethal dose (LD50) for KTR is 189 mg/kg and for AVH is 250 mg/kg. In comparison, the higher the LD50 values (structure of AVH), the less toxic it is compared to KTR. Organ toxicity showed nephrotoxicity of KTR with probability (0.85) active, but for AVH with probability (0.57) inactive. Also, respiratory toxicity showed KTR with probability (0.92) active, but for AVH with probability (0.65) inactive. Also, KTR showed neurotoxicity with probability (0.55) active, but for AVH with probability (0.53) inactive. The results obtained from 45 endpoints and acute toxicity (LD50) show that AVH has significantly fewer toxic effects than KTR. The results of oral toxicity prediction and the toxicity model report for both compounds are presented in detail in Tables S2, S3, and S4. Molecular dynamics (MD) simulations After evaluating the properties of the 20 predicted structures via Spartan, an initial screen was performed using molecular docking. The final selection of the new AVH molecule was then made based on comparison with the KTR reference compound. To investigate the dynamic interactions between the designed AVH molecule and the target protein, classical MD simulations were performed using the GROMACS 2023 suite. These simulations were performed on each target enzyme in complex with the designed molecules. Figure S2 shows the complex @KTR and @AVH in the solvent box. For benchmarking purposes, MD simulations of the KTR drug bound to the protein were also performed. Experimental evidence suggests that integrating protein–ligand binding data with techniques such as MD simulations yields significant insights. While binding provides an estimated spatial arrangement of the ligand in the receptor binding site, careful analysis of other factors, most importantly structural stability and any off-target binding, is essential. The structural stability of the ligand is crucial for effective enzyme inhibition, as it ensures stable occupancy of the binding pocket. MD simulation studies provide extensive analyses of the structural dynamics of protein–ligand complexes under conditions that mimic physiological environments. The MD trajectories obtained from the simulations were used to investigate dynamic equilibrium over time. To assess this equilibrium, the MD trajectories were analyzed, and the root mean square deviation (RMSD) of the simulated structures with respect to the initial structure was calculated. Root mean square deviation (RMSD) RMSD is a common method in molecular dynamics for investigating the stability and structural changes over the simulation time69. In this method, the deviation of the molecular structure at any moment of the simulation from its initial or reference structure is calculated. The value of this parameter helps us to understand the equilibrium and stability of the system over time. Based on this trajectory, as shown in Fig. S3, KTR and AVH remained constant at 10-ns intervals throughout the simulation. To investigate the stability of the ligand–protein binding state during the simulation, the RMSD plot of both AVH and KTR structures was plotted with respect to the protein backbone as the reference structure (Figs. 4 and 5). Two important factors are the fluctuation and trend of the RMSD plot. As can be seen in the plot (Fig. 4), its average variation in the KTR plot is in the range of 1.8–2.7. As the simulation time increased, the fluctuation and trend of the plot became almost stable. As you can see in the AVH plot, the average of these changes is in the range of 0.3–1.1, which is more stable than the KTR structure and indicates a strong and stable binding of this ligand to the protein over the simulation time.Fig. 4RMSD plot of both AVH and KTR structures relative to the protein backbone.Fig. 5RMSD values extracted from protein-KTR and protein-AVH complex for 200 ns. RMSD plot of both AVH and KTR structures relative to the protein backbone. RMSD values extracted from protein-KTR and protein-AVH complex for 200 ns. After analyzing the root mean square deviation (RMSD) of the ligands, we evaluated the RMSD of the entire protein–ligand complex during the simulation. The data presented in Fig. 5 show that both complexes reach a stable conformation by 200 ns, after which minimal deviation is observed. This indicates that the protein structure maintains its integrity during the simulation. In contrast, the RMSD plot for the protein in complex with inhibitors shows significant fluctuations over the same time interval, indicating dynamic motion in the protein structure. It is worth noting that, as mentioned, to ensure the statistical accuracy of the results, we performed three independent 200 ns iterative molecular dynamics simulations for each system. The results from these iterative simulations consistently show that AVH exhibits a completely distinct dynamic stability profile compared to KTR, thereby strengthening our prediction of potential structural substitution. Root mean square fluctuations (RMSF) Root mean square fluctuation (RMSF) is a key technique for analyzing molecular dynamics simulation data70. Essentially, the RMSF value indicates how much an atom or residue deviates from its defined reference point (usually its starting location) along the simulation path. Root mean square fluctuation (RMSF) analysis is used to identify regions in a molecule that exhibit significant dynamic motion or, conversely, regions of high stability. These data are invaluable for understanding the structure–function relationship of biological macromolecules. As shown in Fig. 6, high RMSF values correspond to increased mobility and greater flexibility in that particular region of the molecule. The binding of the AVH ligand caused both increased and decreased flexibility of specific residues, indicating that local structural rearrangements occurred in the protein upon drug binding. Furthermore, the significant decrease in RMSF values, which indicates a decrease in flexibility, may indicate inhibition of aggregation and prevention of insoluble plaque formation. Based on the RMSF profiles, although AVH has a more stable (lower RMSF) behavior than KTR at some residues, it showed a relatively similar behavior at most residues compared to the KTR complex during the simulation, indicating a good similarity between AVH and KTR.Fig. 6RMSF plot showing residue flexibility comparison between KTR and AVH compounds during simulation. RMSF plot showing residue flexibility comparison between KTR and AVH compounds during simulation. Radius of gyration (Rg) The radius of gyration71 (Rg) serves as a measure to quantify the compactness or expansion of a molecule, especially in computational and biophysical studies. Rg represents the average distance of the atoms constituting the molecule from its center of mass. Radius of gyration plots were generated to assess the compactness of ligand–protein complexes formed by two different systems. Analysis of the plots in Fig. S4 reveals that the Rg value for the AVH system decreased after 200 ns, whereas the KTR system exhibited a high Rg. The decrease in Rg indicates a transition towards a more compact and dense conformation over the simulation time frame. In contrast, the significant increase observed for the drug-bound protein suggests that ligand binding causes structural expansion and stabilization. Hydrogen bond and SASA analyses Hydrogen bond analysis in molecular dynamics is a crucial technique for investigating the behavior and properties of chemical and biological systems72. Hydrogen bonds play a critical role in determining the structure, stability, and function of molecules. To investigate the extent to which the solubility and hydrophilicity of a protein change during simulation, the number of hydrogen bonds was plotted separately between the KTR and AVH protein structures as a function of time. As shown in Fig. S5, the increase in the number of hydrogen bonds between the solvent and the protein structure after binding to AVH indicates its increased solubility in water as well as greater interaction with the protein and greater stability of the AVH structure. After 150 ns, the number of hydrogen bonds formed is constant. On the other hand, a further increase in hydrogen bonds may lead to a decrease in the permeability of the drug through biological membranes. The graph shows that this value was favorable in the AVH structure. SAS73 area and volume can reveal changes in protein–ligand structure over time (Fig. 7), reflecting how protein-KTR and AVH adapt. Tracking SAS over time can reveal how proteins unfold or refold during interactions with ligands or under different environmental conditions. Additionally, changes in SAS can correlate with structural stability indicators, such as the radius of gyration (Rg), which helps assess whether a protein maintains its compactness or becomes more extended during simulations.Fig. 7Time evolution of solvent-accessible surface (SAS) (a) area and (b) volume for the complexes formed with KTR and AVH. Time evolution of solvent-accessible surface (SAS) (a) area and (b) volume for the complexes formed with KTR and AVH. Coulombic and Lennard–Jones interaction energies at short distances The potential energy arising from the forces between individual components within a molecular system is known as interaction energy. Analyzing interaction energies offers valuable insights into the characteristics of a molecular system74. For instance, the stability of a molecular conformation can be assessed by monitoring fluctuations in interaction energy. Furthermore, comparing interaction energies between distinct molecules reveals their relative binding affinities. In the context of Coulombic forces, stronger electrostatic attractions with the target protein may suggest an enhanced drug binding capacity and increased stability (Fig. S6), as demonstrated in the accompanying graph. Additionally, more favorable Lennard–Jones interactions, indicative of optimal van der Waals potential energy, can signify improved drug compatibility with its surrounding environment and neighboring molecules. Notably, the AVH drug (Fig. S7) structure exhibited an even higher value in this regard compared to KTR. DSSP DSSP analyzes secondary structure assignments by examining a protein’s secondary structure; however, it is not a predictive method75. The algorithm recognizes structural motifs, such as helical turns and beta bridges, according to defined hydrogen bonding geometries. DSSP analysis is capable of classifying all protein structural features, including loops, alpha helices, coils, and beta sheets. As illustrated in Fig. 8, the two complexes exhibit remarkably similar conformational shifts. The alterations observed in the beta sheet and alpha helix structures are virtually identical. Furthermore, the inclusion of AVH appears to reduce the bend in the structure and enhance the stability of the beta sheet.Fig. 8The secondary structure of the protein after adsorption of (a) KTR and (b) AVH, showing conformational changes induced by each ligand. The secondary structure of the protein after adsorption of (a) KTR and (b) AVH, showing conformational changes induced by each ligand. These findings indicate that the AVH compound, as shown in Fig. 8, does not induce significant structural changes in the protein. Its behavior is exactly similar to that of KTR and, in some aspects, even surpasses its function. The secondary structure heatmap and the secondary structure fraction per residue76 show how the secondary structure at each specific residue position changes over time. The AVH and KTR secondary structure heat maps are shown in Fig. S8a,c, which show the temporal evolution of the secondary structure of a protein over approximately 80,000 frames. The vertical axis is labeled with 6-residue intervals (1–127) and shows a color bar: purple (coil), cyan (strand), yellow (helix). It also shows that a stable helical cluster is formed in the region 18–32 and is maintained throughout the simulation. This allows for the visualization of dynamic changes in helices, strands, or coils that occur at specific residues during the simulation. Overall, these heatmaps provide a detailed temporal and spatial view of the secondary structure dynamics in the protein during the simulation. Figure S8b,d also show the secondary structure fractions per residue for both of the aforementioned conformations, respectively. These bar graphs quantify the fraction of simulation frames that adopt each secondary structure class per residue. The bottom panel shows the same data side by side for easier comparison. A residue with a predominantly red bar means that it remains helical most of the time; predominantly blue means that it prefers a single strand; green means that it is helical. This quantifies the trends visible in the heatmap, such as which residues have stable secondary structures or are more flexible. Together, these plots provide a dynamic map of the protein’s secondary structure stability during the simulation. Dynamic cross-correlation matrix (DCCM) analysis Dynamic cross-correlation matrix (DCCM)77 analysis is a computational technique used in molecular dynamics (MD) simulations to quantify the degree of correlated motion between atoms or residues within a molecule, such as a protein. This analysis generates a matrix that measures the interrelationship between the movements of different atoms or residues throughout the simulation. Figure S9 presents the backbone displacement cross-correlation matrix, highlighting both correlated and uncorrelated motions in the two conformations. The diagonal regions exhibit pronounced correlated motions. The matrix elements are symmetric about the diagonal. In the correlation maps, red areas represent strongly correlated residue motions, whereas blue areas indicate strongly anti-correlated motions. Comparison of the KTR and AVH system maps shows the results of correlated movements. (a) Shows a more broadly distributed pattern of correlations and anticorrelations, with significant blocks of off-diagonal red and blue elements, indicating both positive and negative correlated motions among residue groups. In contrast, (b) shows a stronger and more continuous band of positive correlations along the main diameter with extended off-diagonal red clusters that are more consistent and concentrated with increased long-range coupling and more coherent global motions along distant residues, indicating a relatively more coordinated motion. Furthermore, (a) shows a protein with complex dynamical behavior, which may be associated with a flexible state. In contrast, the more concentrated correlation pattern in (b) indicates a more rigid or structurally stable state, likely after ligand binding, mutation, or under different environmental conditions that reduce dynamical variability. Free energy landscape, principal component, and porcupine plot analysis Principal component analysis based on a free energy landscape is a central method in molecular dynamics simulation. It reduces the complexity of molecular simulation data and interprets the energy governing the function and dynamics of biomolecular systems78. To characterize the apparent structural dynamics of KTR and AVH, we performed a comprehensive analysis that combined principal component analysis (PCA) with free energy landscape (FEL) visualization via contour plots. In these plots, the longitudinal axis and the longitudinal axis correspond to the first (PC1) and second (PC2) principal components, respectively, while the accompanying color gradient encodes the structural free energy. The principal components capture the dominant collective motions of the proteins, with PC1 and PC2 representing the main bidirectional oscillations evaluated here. This approach enabled the identification of predominant conformational states and transition pathways throughout the simulation. As shown in Fig. S10, the 2D and 3D results indicate that AVH forms a more stable protein–ligand complex upon binding, as evidenced by a well-defined cluster with minimal exploration of the structural phase space compared to KTR. Porcupine plots79, which are drawn using the custom PyMOL43 program, highlight collective motions along a principal coordinate commonly used to visualize the results of principal component analysis (PCA) in molecular dynamics simulations, and provide insight into how ligand binding affects protein dynamics. Longer, directional arrows identify “hot spots” of flexibility, while shorter arrows indicate significant stabilization in those regions. Figure S11a shows prominent arrows on surface loops and specific regions where drug binding in this scenario causes directional motions in those regions. In contrast, Figure (b) shows distributed red arrows where AVH causes significant motion and likely increases stability in other regions. Electrostatic surface potential calculation To elucidate the charge distribution and characterize the interaction dynamics, electrostatic surface potential mapping was performed using PyMOL80. Electrostatic potential analysis is a powerful computational approach for delineating the spatial distribution of surface charges on proteins and inferring their binding affinity and specificity toward ligands. As shown in Fig. S12, diagram of the potential (a) KTR and (b) AVH in the active site of the protein, and in Fig. 9, the electrostatic potential map of (a) KTR and (b) AVH, the protein binding cavity shows negative and neutral electrostatic regions that probably contribute to further stabilization of AVH in the active site.Fig. 9Surface electrostatic potential map protein (a) KTR and (b) AVH calculated by PyMOL software. The positively charged region is shown with blue patches and the negatively charged region is shown with red patches. Surface electrostatic potential map protein (a) KTR and (b) AVH calculated by PyMOL software. The positively charged region is shown with blue patches and the negatively charged region is shown with red patches. MM-PBSA analyses MM-PBSA is a computational technique that combines molecular mechanics with continuum solvation models to estimate the binding free energies of biological molecules, such as proteins and ligands. The method calculates the internal energy using molecular mechanics, encompassing both bonding and non-bonding interactions. The Poisson-Boltzmann equation models the effects of electrostatic solvation, and the surface area term approximates the contribution of nonpolar solvation using the solvent-accessible surface area. The MMPBSA method was used to calculate the binding free energy and investigate the interactions between a target residue and a ligand substructure. A negative binding energy indicates that the binding is spontaneous. As shown in Fig. 10, although the binding energy of AVH (− 16.96 kcal/mol) is slightly higher than that of KTR (− 23.65 kcal/mol), it is still a negative binding energy, which is thermodynamically favorable. Table S1 reports the numerical values of the energy parameters.Fig. 10MM-PBSA energetic components and their values. ΔEvdW: Van der Waals energy contribution; ΔEele: Electrostatic interactions in a vacuum; ΔEGB: the electrostatic contribution to the solvation free energy calculated by PB; ΔESURF: Nonpolar solvation energy calculated by an empirical model; ΔGGAS: Total gas-phase energy (ΔEvdW + ΔEele); ΔGSOLV: Total solvation energy (ΔEGB + ΔESURF) for (a) KTR and (b) AVH. MM-PBSA energetic components and their values. ΔEvdW: Van der Waals energy contribution; ΔEele: Electrostatic interactions in a vacuum; ΔEGB: the electrostatic contribution to the solvation free energy calculated by PB; ΔESURF: Nonpolar solvation energy calculated by an empirical model; ΔGGAS: Total gas-phase energy (ΔEvdW + ΔEele); ΔGSOLV: Total solvation energy (ΔEGB + ΔESURF) for (a) KTR and (b) AVH. Experimental Materials and Methods L-valine, allyl alcohol, ninhydrin, Diethyl ether, chloroform (CHCl3), glacial acetic acid, n-butyl alcohol, acetone, aluminum TLC plate Silica Gel 60 F254, CDCl3, and Triethylamine (NEt3) were all purchased from Merck. Trimethylsilyl chloride (TMSCl) was purchased from Aldrich and distilled before use. FTIR spectra of the samples were recorded using an ABB Bomem MB-100 FTIR spectrophotometer. Nuclear magnetic resonance (NMR) spectra were recorded. General procedure for the preparation of valine allyl ester hydrochloride (AVH) According to Fig. S13, initially (0.352 g, 0.003 mol), valine was added to a round-bottom flask. Then, 15 mL of allyl alcohol was added to it while stirring at room temperature. Thirty minutes later, freshly distilled Trimethylsilyl chloride (0.89 mL, 0.007 mol) was added dropwise. When the injection was complete, we raised the reaction temperature to 50 °C. TLC monitored the progress of the reaction. TLC development of n-butyl alcohol and distilled water was mixed. Glacial acetic acid was then added to this mixture, shaken well, and left undisturbed for 15 min to allow the layers to separate. The upper layer was carefully separated and used as the mobile phase. Prepare a 0.5% solution of ninhydrin in a mixture of acetone and water and spray it onto the TLC paper. Then, heat the paper with a hair dryer for 2 min. Afterward, the spots will become visible. After completion, the reaction mixture was neutralized with Et3N. Then, diethyl ether was added. The crude precipitate was filtered. 10 cc of CHCl3 was added and stirred for one hour (the product is soluble in CHCl3, but valine is insoluble). Then, it was filtered, and diethyl ether was added to the solution under the filter. The brown precipitate formed was filtered and dried in vacuo. The FT-IR spectrum (Fig. S14) and 1H NMR and 13C NMR spectra of the product are presented in Figs. S15 and S16. Conclusion In this study, we investigated the potential of l-amino acids as a basis for designing new drugs with fewer side effects, with a particular focus on finding alternatives to ketorolac. Using computer-aided drug discovery methods, we provided a detailed atomic-level description of the drug ligand and evaluated its selectivity properties, ADMET profiles, and Pre-Tox analysis. Through molecular docking calculations, we identified the optimal active site for the new drugs, which is similar to the protein localization of ketorolac. Among the proposed structures, AVH showed the most promise, which was very similar to ketorolac. Molecular dynamics simulations revealed the stability and strong interaction between the protein and the designed AVH ligand, as confirmed by RMSD, RMSF, gyration, and hydrogen bonding plots. The DSSP algorithm further confirmed the suitability of the proposed AVH structure by demonstrating that the protein’s secondary structure and residue-based secondary structure remained intact, and in some cases, performed better than ketorolac. DCCM analysis revealed a stronger correlation pattern in AVH, indicating greater stability in this pattern. Principal component analysis, based on a free energy perspective, revealed closely spaced data points representing low-energy and relatively stable states that are energetically favorable for the system. Also, Porcupine plots show distributed red arrows where AVH causes significant motion and likely increases stability in other regions. The protein binding cavity observed in the electrostatic potential map likely indicates greater stabilization of the ligand in the active site. Finally, the MM-PBSA results show that although the binding energy of AVH is slightly higher than that of KTR, it remains negative, indicating a favorable thermodynamic binding and a strong affinity of AVH for the protein. After molecular docking, MD simulations, ADMET prediction, Pre-Tox analysis, and MM-PBSA evaluation, the AVH structure was synthesized. The results of this research indicate that the new l-amino acid-based structure, designed based on a natural product with similar chemical effects, is a promising potential alternative to ketorolac. This study provides valuable insights into the development of safer and more effective drugs based on l-amino acids. It paves the way for future research and applications in the pharmaceutical field81. Supplementary Information Supplementary Information. Supplementary Information.