Recep Adiyaman
bioinformatics

Issue #3: A Comparative Study of Deep Learning and Classical Modeling Approaches for Protein-Ligand Binding Pose and Affinity Prediction in Coronavirus Main Proteases.

December 23, 2025 Daily Intelligence
Protein Design Daily

Building something in Protein Design?

I love collaborating on new architectural challenges. Let's build together.

LinkedIn Connect Direct Message

🧬 Protein Design Digest

Curated protein signals by Recep Adiyaman

Join 1,000+ researchers. Unsubscribe anytime.

🚀 Today’s Top Signal

A Comparative Study of Deep Learning and Classical Modeling Approaches for Protein-Ligand Binding Pose and Affinity Prediction in Coronavirus Main Proteases.

🧬 Abstract

The accurate prediction of protein-ligand binding poses and affinities is central to structure-based drug design. In this study, we first benchmarked three distinct pose generation strategies for data sets from the ASAP Antiviral Challenge 2025: molecular docking (Glide and AutoDock Vina), ligand-based superposition (FlexS), and deep learning-based modeling (AlphaFold3, Boltz-2, DiffDock and Gnina). We evaluated their performance on binding pose prediction for ligands targeting SARS-CoV-2 and MERS-CoV main protease (Mpro). For binding affinity estimation, we implemented a machine learning-based scoring approach called ligand-residue interaction profile scoring function (LRIP-SF), which integrates molecular mechanics generalized Born surface area (MM-GBSA) energy decomposition with machine learning algorithms. Our results showed that deep learning-based modeling with AlphaFold3 achieved the highest pose prediction accuracy with a success rate of 88.1% and an average ligand root-mean-square deviation (LRMSD) of 1.12 Å. Moreover, binding poses predicted by AlphaFold3 enabled the most accurate potency predictions by LRIP-SF, with the lowest mean absolute error (MAE) and root-mean-square error (RMSE) in pIC50 units across both targets: the MAE and RMSE are 0.606 and 0.813, respectively, for MERS-CoV Mpro and 0.724 and 0.894 respectively for SARS-CoV-2 Mpro. Although ligand-based superposition method (FlexS) was less accurate in pose prediction, it offered competitive potency prediction performance with significantly lower computational cost. To interpret model predictions by LRIP-SF and identify critical binding determinants, we performed global sensitivity analysis (GSA), revealing key residues that contributed most significantly to ligand binding. These findings highlight the importance of pose quality and interaction profiling in affinity prediction and demonstrate the great potential of deep learning-based methods for drug discovery, especially in the absence of cocrystal structures.

Why it matters: Critical for improving fold accuracy and reducing structural uncertainty in de novo design.

⭐ Additional Signals

Cytotoxicity, apoptosis, molecular docking, and molecular dynamics study of novel compounds of Sulfamide derivatives coupled with DHP scaffolds as potent inhibitors of the MCF-7, A549, SKOV-3, and EA. yh926 carcinoma cells.

A novel series of dihydropyridine-sulfonyl derivatives (AG-CHO and analogues A1-A7) were synthesized and structurally characterized. Molecular docking demonstrated favorable binding of these compounds to autophagy-associated and cancer-related targets, while molecular dynamics simulations confirmed A5 as the most stable ligand protein interactions. Functional assays in SKOV-3, MCF-7, A549, and EA.hy.926 cells using acridine orange staining and flow cytometry revealed significant autophagy induction. Among all tested compounds AG-CHO emerged as the most potent inducer of autophagy. Notably, derivatives such as A6 and A7 showed selective potency in endothelial cells, whereas A1, A5, and A7 were effective in A549 cells, indicating cell-specific activity. Collectively, this integrated computational and experimental study identifies A5 as the lead compound and highlights dihydropyridine-sulfonyl scaffolds as promising autophagy modulators and potential anticancer candidates for further preclinical development.

Meeko: Molecule Parametrization and Software Interoperability for Docking and Beyond.

Molecule parametrization is an essential requirement to guarantee the accuracy of docking calculations. Parametrization includes a proper perception of chemical properties such as bonds, formal charges and protonation states. This includes large biological macromolecules, such as proteins and nucleic acids, and small molecules, such as ligands and cofactors. The structures of proteins and nucleic acids are challenging due to omission of several atoms from the structural model, and from the lack of connectivity and bond order information in the PDB and mmCIF file formats. For small molecules, the very large chemical diversity poses challenges for both validating correctness and providing accurate parameters. These challenges affect various modeling approaches like molecular docking and molecular dynamics. Moreover, several specialized methods (particularly in molecular docking) leverage specific chemical properties to add custom potentials, pseudoatoms, or manipulate atomic connectivity. To address these challenges, we developed Meeko, a molecular parametrization Python package that leverages the widely used RDKit cheminformatics library for a chemically accurate description of the molecular representation. Small molecules are modeled as single RDKit molecules, and biological macromolecules as multiple RDKit molecules, one for each residue. Meeko is highly customizable and designed to be easily scriptable for high-throughput processing, replacing MGLTools for receptor and ligand preparation.

🧪 AI & Research News

🏢 Industry Insight & Applications

⚡ Quick Reads

A fully automated benchmarking suite to compare macromolecular complexes.

Protein structure prediction has a long history of benchmarking efforts such as critical assessment of structure prediction, continuous automated model evaluation and critical assessment of prediction of interactions. With the rise of artificial intelligence-based methods for prediction of macromolecular complexes, benchmarking with large datasets and robust, unsupervised scores to compare predictions against a reference has become essential. Also, the increasing size and complexity of experimentally determined reference structures by crystallography or cryogenic electron microscopy poses challenges for structure comparison methods. Here we review the current state of the art in scoring methodologies, identify existing limitations and present more suitable approaches for scoring of tertiary and quaternary structures, protein-protein interfaces and protein-ligand complexes. Our methods are designed to scale efficiently, enabling the assessment of large, complex systems. All developments are available in the structure benchmarking framework of OpenStructure. OpenStructure is open source software and available for free at https://openstructure.org/ .

From sweetener to risk factor: Network toxicology, molecular docking and molecular dynamics reveal the mechanism of aspartame in promoting coronary heart disease.

Aspartame, a widely used non-nutritive sweetener, has been epidemiologically linked to coronary heart disease (CHD), although the underlying mechanisms remain unclear. This study employed an integrative computational strategy combining network toxicology, molecular docking, and molecular dynamics to decode aspartame’s CHD-promoting mechanisms. Initially, the toxicity profile of aspartame was predicted using ProTox 3.0 and ADMETlab 3.0, which highlighted significant cardiotoxicity. Through multi-source target screening of aspartame (PharmMapper, SEA, etc.) and CHD (GeneCards, OMIM), 216 shared targets were identified. Protein-protein interaction network analysis revealed 10 hub targets (INS, PPARGC1A, TNF, AKT1, IL6, MMP9, IGF1, PTGS2, SIRT1, PPARG). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses revealed significant enrichment in lipid metabolism, inflammatory responses, insulin resistance, and atherosclerosis-related pathways. Molecular docking and molecular dynamics simulations (MDS) demonstrated high-affinity binding of aspartame to three core targets (PTGS2, TNF, and PPARGC1A), with a binding energy ≤ -7.0 kcal/mol, and confirmed high binding stability. This study reveals that aspartame may promote the pathogenesis of CHD by disrupting cardiovascular homeostasis through multi-target interactions, including inflammatory response, metabolic dysregulation, and vascular remodeling. These findings provide molecular evidence for re-evaluating the safety profile of aspartame and establish a computational framework to guide experimental validation and preventive strategies.

Multi-target exploration of newly synthesized pyrazoline-quinoline derivatives via in vitro screening, QSAR, molecular docking, MD simulations, and DFT analysis.

The development of multifunctional therapeutic agents remains a promising strategy in modern drug discovery, particularly for diseases associated with oxidative stress, bacterial infections, and cancer progression. In this study, a new series of [5-(substituted phenyl)-3-(substituted phenyl)-4,5-dihydro-pyrazol-1-yl]-(2-methyl-quinolin-4-yl)-methanones (9a-o) has been synthesized and evaluated for anticancer, antibacterial, and antioxidant activities through established in vitro and in silico screening models. The in vitro cytotoxic evaluation conducted against the human lung cancer cell line (A549) using the MTT assay revealed that all synthesized compounds have significant inhibitory potential. Among them, compounds 9i, 9b, 9h, 9d, and 9o demonstrated superior potency, showing IC₅₀ values of 3.68 ± 0.45 μM, 4.06 ± 0.35 μM, 4.33 ± 0.68 μM, 6.32 ± 0.89 μM, and 7.82 ± 0.52 μM compared to the reference drug doxorubicin, which showed an IC₅₀ of 9.48 ± 0.35 μM under identical experimental conditions. The same compounds also possess the best antibacterial (MIC value 12.5-25 μg/mL) and antioxidant potential. The in silico studies encompassed ADMET analysis, QSAR, molecular docking, and molecular dynamics simulations, which were carried out using Molsoft LLC, pkCSM, ChemDes, AutoDock 4.2, and GROMACS software. Their electronic and reactivity features were also analyzed through DFT calculations based on HOMO, LUMO, electron affinity, ionization potential, chemical potential, and global softness. The results of computational studies reinforced the findings in all dimensions. In summary, this study introduces a promising class of pyrazoline-quinoline conjugates with significant multifunctional efficacy.

UPLC-Q-TOF/MS-based Spectrum-effect Correlation Combined with Chemometrics and Molecular Docking for Quality Assessment and Screening of Bioactive Components with Hemostatic, Antinociceptive, and Anti-Inflammatory Activities in Liparis nervosa.

Ethnopharmacological relevance Liparis nervosa (LN) occurs in Southwest China and is traditionally used as a hemostatic and detoxifying agent; however, the pharmacodynamic basis for its medicinal properties is unclear; this impedes the quality standardization and clinical application of this herb. Aim of the study This study aimed to establish an integrated quality assessment system for LN by combining comprehensive chemical profiling with pharmacological evaluation to identify bioactive components and quality markers. Materials and methods Chemical profiling of ten regional LN specimens via UPLC-Q-TOF/MS revealed 53 shared components and characteristic fingerprints. Concurrently, systematic evaluation of hemostatic, antinociceptive, and anti-inflammatory activities was used to identify bioactive fractions. Using spectrum-effect modeling, which integrates techniques such as gray relational analysis, partial least squares regression, and bivariate correlation linked chromatographic features to bioactivities, these pharmacological effects were correlated with specific chemical components. Molecular docking was performed to validate target interactions. Orthogonal design coupled with spectrum-effect relationship analysis was used to pinpoint potential quality markers. Results As the first comprehensive study to systematically identify bioactive fractions and quality markers of LN, this work developed a tripartite evaluation framework integrating chemical profiling, pharmacological verification, and molecular docking-based target validation. Conclusions This methodology advances the standardization of LN, supports the interpretation of its pharmacological mechanisms of action, and facilitates the development of multi-target phytotherapeutic agents using LN bioactives.

Exploring the Mechanism of Platycladi Cacumen in Intervening Androgenetic Alopecia Based on Network Pharmacology, Molecular Docking, and Molecular Dynamics Simulation

Abstract As a traditional hair-growth-promoting herb, Platycladi Cacumen(PC) has a long history of folk application in the field of hair loss improvement. Preliminary modern pharmacological studies have suggested that its active components may exert potential effects by regulating hair follicle-related signaling pathways; however, for androgenetic alopecia (AGA), the exact targets and specific regulatory mechanisms of PC remain unelucidated, which provides a direction for research on natural drug-based intervention in AGA. In this study, network pharmacology was employed to predict the active components and core targets of PC. Targets associated with AGA were collected, and the intersection targets between PC and AGA were identified. Subsequently, protein-protein interaction (PPI) analysis, Gene Ontology (GO) enrichment analysis, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed on the intersection targets to screen out the core targets. Thereafter, molecular docking and molecular dynamics simulation were conducted to validate the interactions between key active components and core targets. The component-target network diagram included 1044 interaction relationships between 32 components and 439 targets, among which quercetin, apigenin, myricetin, and hinokinin were identified as key components. The disease-target network diagram summarized 410 targets associated with AGA. Through PPI network analysis, key targets such as ESR1, BCL2, INS, AR, and STAT3 were screened out. The results of GO enrichment analysis and KEGG pathway analysis revealed that PC may exert its effects by regulating the EGFR receptor molecule and pathways including the HIF-1 signaling pathway. Molecular docking results showed that the binding energies of all complexes were less than -6.4 kcal/mol, indicating favorable binding effects. Molecular dynamics simulation results showed that the root-mean-square deviation (RMSD), root-mean-square fluctuation (RMSF), radius of gyration (Rg), solvent-accessible surface area (SASA), two-dimensional free energy landscape (FEL-2D), and FEL-3D of the simulation system all remained in an equilibrium state with small fluctuation amplitudes. This result indicated that the molecular system had a stable overall conformation, restricted local residue movement, a compact spatial structure, and stable internal chemical bonds—collectively confirming that the quercetin-STAT3, apigenin-AR, myricetin-STAT3, and hinokinin-AR complexes exhibited extremely strong binding stability. Collectively, Overall, this study systematically investigated the mechanism of action and potential value of PC leaves in intervening in AGA, providing a solid theoretical basis for the intervention of AGA with PC.

A hardware demonstration of a universal programmable RRAM-based probabilistic computer for molecular docking.

Molecular docking is a critical computational strategy in drug discovery, but the diversity of biomolecular structures and flexible binding conformations create an enormous search space that challenges conventional computing. Quantum computing holds promise but remains constrained by scalability, hardware limitations, and precision issues. Here, we report a probabilistic computer (p-computer) prototype that solves complex molecular docking. The system is built upon artificial probabilistic bits (p-bits), fabricated in 180 nm CMOS with BEOL HfO₂ RRAM and compatible with compute-in-memory (CIM) schemes. A key innovation is the integration of Gaussian Random Number Generator-based p-bits with CIM, where the sigmoidal response arises from the Gaussian cumulative distribution function with coupling and bias coefficients directly encoded in the RRAM crossbar. This co-design alleviates the memory-to-compute bottleneck of prior CMOS-only and CMOS + X (emerging nanodevices) p-computers. Using this architecture, we experimentally solved a 42-node docking problem of lipoprotein with the LolA-LolCDE complex-a key target in developing antibiotics against Gram-negative bacteria, with results consistent with the Protein-Ligand Interaction Profiler tool. This work represents an early hardware application of p-computing in computational biology and demonstrates its potential to overcome the success rate and efficiency limitations of current technologies for complex bioinformatics problems.

💡 Pipeline Tip

Check for missing residues in PDB files using PDB-Fixer before simulation.

🛠️ Resources

The protein structure is the language of life; design is its poetry. — Recep Adiyaman

BS HF DK