📄 chemistry

In silico design of potent antioxidant compounds using 2D-QSAR Molecular docking ADMET prediction and Molecular dynamic simulation

This study employed an integrated in silico approach combining 2D-QSAR modeling, molecular docking, dynamic simulations, and ADMET predictions to design and evaluate novel benzoxazine derivatives as potent and drug-like antioxidant agents targeting cytochrome c peroxidase.

Original authors: Muhammad Baba Muhammad, Adamu Uzairu, Mohamed EL-fadili, Stephen Eyije Abechi, Muhammad Tukur Ibrahim, Somdutt Mujwar

Published 2026-07-02

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Muhammad Baba Muhammad, Adamu Uzairu, Mohamed EL-fadili, Stephen Eyije Abechi, Muhammad Tukur Ibrahim, Somdutt Mujwar

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Fighting the "Rust" in Our Bodies

Imagine your body is a high-tech car. Over time, just like a car, it gets "rusty." In biology, this rust is called oxidative stress, caused by tiny, chaotic particles called free radicals (or Reactive Oxygen Species). These particles are like little sparks flying around inside your engine. If there are too many of them, they start burning up your fuel lines (cell membranes) and snapping your wires (DNA). This leads to "wear and tear" diseases like heart trouble, diabetes, and Alzheimer's.

To stop the rust, your body has a built-in mechanic called Cytochrome c peroxidase. Its job is to catch those sparks and neutralize them. However, sometimes the mechanic gets overwhelmed. This study is about designing a new, super-powered "assistant mechanic" (a drug) to help out.

The Strategy: The Digital Workshop

Instead of mixing chemicals in a messy lab and waiting years to see what works, the researchers used a digital workshop (computers) to design and test these new assistants. They used four main tools:

The Pattern Matcher (QSAR): They looked at 28 existing chemical shapes (benzoxazines) and asked, "What makes the best ones work?" It's like analyzing 28 different keys to find out which teeth on the key open the lock best. They built a mathematical rulebook to predict which new shapes would be the best keys.
The Virtual Locksmith (Molecular Docking): They took their best new designs and tried to fit them into the "lock" (the Cytochrome c peroxidase enzyme) on a computer screen. They wanted to see if the new keys fit tighter and turned the lock better than the old ones.
The Stress Test (Molecular Dynamics): Just because a key fits once doesn't mean it stays there. They ran a 100-second "movie" (simulation) to see if the key wiggled loose or stayed firmly locked in place while the engine vibrated.
The Safety Inspector (ADMET): Before a car leaves the factory, it must pass safety checks. The researchers checked if their new keys would dissolve in the stomach, travel through the blood, or accidentally poison the brain.

What They Found

1. The Best "Keys" (The New Designs)
The researchers started with a "template" compound (Compound 20) that was already pretty good. Using their pattern-matching rulebook, they tweaked the template by adding small chemical "handles" (like hydroxyl or methoxy groups) to make it fit better.

The Result: They created five new designs (labeled D1 through D5).
The Score: In the virtual lock-picking test, the new designs scored much higher than the original template and even better than Vitamin C (the standard antioxidant everyone knows). Specifically, design D5 was the strongest, followed closely by D1 and D4. They gripped the enzyme so tightly that the computer calculated a very strong "hug" (binding energy).

2. The Stability Check
The researchers watched the best designs (D1 and D5) for a long time in the simulation.

The Result: They stayed put. Even though the protein wiggled around, the new keys didn't fall out. They found a comfortable spot and held on tight, proving they are stable enough to do their job.

3. The Safety Report
Before a drug can be real, it has to be safe.

Absorption: The computer predicted these new compounds would be easily absorbed by the human intestine (like a sponge soaking up water), meaning if you took them as a pill, your body would actually use them.
Brain Safety: They are too big or the wrong shape to cross the "brain barrier." This is actually good news because it means they won't accidentally cause confusion or damage to the brain.
Toxicity: Most of the new designs passed the "mutagenicity" test (meaning they won't likely cause cancer or genetic damage), though a couple of them (D3 and D4) showed a slight risk of being mutagenic.

The Bottom Line

The researchers successfully used a computer to design five new chemical shapes based on the benzoxazine family.

They fit the target enzyme better than Vitamin C.
They stay stuck to the target without falling off.
They look safe to travel through the human body.

The Catch: The paper explicitly states that these are computer predictions. The authors conclude that while these compounds look promising and "deserve a try," they have not been tested in real living cells (in vitro) or in living animals/humans (in vivo) yet. The study is a blueprint for a better car part, but the part hasn't been installed in a real engine yet.

Technical Summary: In Silico Design of Potent Antioxidant Compounds

Problem Statement
Oxidative stress, driven by an imbalance between reactive oxygen species (ROS) and endogenous antioxidant defenses, is a primary factor in aging and degenerative disorders such as cancer, cardiovascular disease, diabetes, and Alzheimer's. While natural and synthetic antioxidants exist, many suffer from limitations including poor solubility, low abundance, and toxicity. Specifically, 1,4-benzoxazine derivatives have shown potential as antioxidants, yet there is a lack of comprehensive computational evaluations linking their structural properties to biological activity, binding modes, and pharmacokinetic profiles. This study addresses the need to rationally design more potent, stable, and safe benzoxazine-based antioxidants using a multi-step in silico workflow.

Methodology
The research employed a ligand-based and structure-based drug design approach involving the following sequential steps:

Dataset Preparation: A dataset of 28 known 1,4-benzoxazine derivatives with reported DPPH radical scavenging activities (IC50) was compiled. Structures were converted to 3D, energy-minimized using Molecular Mechanics (MMFF), and optimized via Density Functional Theory (DFT) at the B3LYP/6-31G* level.
2D-QSAR Modeling: Molecular descriptors were generated using PADEL-Descriptor. The dataset was split into training (75%) and test (25%) sets using the Kennard-Stone algorithm. Four models were built using Genetic Function Algorithm (GFA) and Multi-Linear Regression (MLR) in Material Studio v8.0. Model validation included internal ( $R^2_{trn}$ , $Q^2_{LOO}$ ), external ( $R^2_{ext}$ ), and Y-randomization ( $cR^2_p$ ) tests. The applicability domain was assessed via the Williams plot.
Molecular Docking: The best QSAR model was used to guide the design of new analogues. All 28 original compounds and five newly designed derivatives were docked against Cytochrome c Peroxidase (PDB ID: 2XO8), a key enzyme in mitochondrial electron transport and apoptosis regulation, using Molegro Virtual Docker (MVD). Ascorbic acid served as the reference standard.
Ligand-Based Design: Compound 20 (the most active template, pIC50 = 7.817) was modified by adding functional groups (hydroxyl, methoxy, ethoxy, ethyl, chlorate, aldehyde) at specific positions (D, X, Y) based on the descriptor contributions in the QSAR model.
ADMET and Pharmacokinetics: Drug-likeness (Lipinski's Rule of 5, Veber's rule) and ADMET properties (absorption, distribution, metabolism, excretion, toxicity) were predicted using SwissADME and pkCSM servers.
Molecular Dynamics (MD) Simulation: The stability of the top-ranked designed complexes (D1 and D5) with the receptor was validated over a 100 ns simulation using Schrödinger's Desmond software. Parameters analyzed included RMSD, RMSF, Radius of Gyration (Rg), and hydrogen bonding.

Key Results

QSAR Model Performance: Model 4 was selected as the optimal model based on statistical robustness ( $R^2_{trn} = 0.7889$ , $R^2_{ext} = 0.7449$ , $Q^2_{LOO} = 0.6889$ , $cR^2_p = 0.6015$ ). The model utilized descriptors GATS2m, GATS3m, SpMin6_Bhv, and IC2. The Y-randomization test confirmed the model was not a chance correlation.
Designed Compounds: Five new benzoxazine derivatives (D1–D5) were designed. They exhibited superior predicted antioxidant activities with pIC50 values ranging from 8.002 to 8.532, outperforming the template (Compound 20, pIC50 = 7.817) and ascorbic acid (pIC50 = 7.817).
Docking Affinity: The designed compounds showed strong binding affinities. Compound D5 achieved the highest MolDock score (-135.37 kcal/mol), followed by D1 (-134.38 kcal/mol) and D4 (-131.14 kcal/mol). These scores were significantly better than the reference ascorbic acid (-82.772 kcal/mol). Interaction analysis revealed diverse binding modes involving hydrogen bonds, pi-cation, pi-pi stacking, and hydrophobic interactions with residues such as ARG 48, TRP 51, and HIS 181.
ADMET Profiles: All designed compounds adhered to Lipinski's and Veber's rules, indicating good oral bioavailability. They demonstrated high gastrointestinal absorption (90–94%), low blood-brain barrier permeability (log BB < 0.3), and favorable metabolic profiles (CYP3A4 substrates, non-inhibitors of CYP2D6). Toxicity predictions indicated that D1, D2, and D5 were non-mutagenic, while D3 and D4 were predicted to be mutagenic.
MD Simulation: The 100 ns MD simulations confirmed the structural stability of the D1 and D5 complexes. The protein backbone RMSD stabilized between 2.2–2.6 Å, and the ligands remained stably bound within the active site, with RMSD values stabilizing around 0.9 Å. Hydrogen bond analysis and contact maps indicated sustained and robust interactions throughout the simulation.

Significance and Claims
The authors claim that this study successfully demonstrates the utility of an integrated in silico pipeline (QSAR, docking, MD, and ADMET) for the rational design of novel antioxidant agents. The study identifies specific structural modifications to the 1,4-benzoxazine scaffold that enhance antioxidant potency and binding affinity to Cytochrome c Peroxidase.

The paper concludes that the designed benzoxazine derivatives, particularly D1, D2, and D5, represent promising candidates for further development as therapeutic agents against oxidative stress-related disorders. However, the authors maintain a modest stance, explicitly stating that these findings are computational predictions that "may serve as promising antioxidant agents subject to further in vitro and in vivo experiments." The work aims to bridge the gap between experimental research and practical therapeutic application by providing a validated structural framework for future synthesis and biological testing.

The Big Picture: Fighting the "Rust" in Our Bodies

The Strategy: The Digital Workshop

What They Found

The Bottom Line

More like this