Structure-based Methods Research Articles

Novel Dipeptide Inhibitors of PfPNP: In-Silico Identification of Promising New Antimalarials.

Malaria, an infectious disease caused by Plasmodium falciparum, is becoming increasingly difficult to treat due to the emergence of drug-resistant strains. Recent studies have proposed purine nucleoside phosphorylase from P. falciparum (PfPNP) as a potential target for malaria treatment. In the present study, we designed a virtual library of 400 dipeptides to discover novel anti-malarial peptide inhibitors. A structure-based molecular docking method was employed to virtually screen the designed library against the wild-type structure of PfPNP (PDB: 5ZNC). The best four (Phe-Arg, Arg-His, Trp-Arg and Tyr-Arg) dipeptides, which were then investigated for their binding potential against PfPNP using Molecular Dynamics simulation studies. Parameters such as RMSD, RMSF, Rg, and SASA were analyzed to understand the structural changes, energetics, and overall behavior of PfPNP-dipeptide complexes. The PfPNP demonstrated significant stability upon binding with each of the identified dipeptides with ΔG of over -168 kcal/mol. Additionally, DFT and ADME predictions indicated that the electronic structure, energetics, and pharmacokinetic properties of Phe-Arg, Arg-His, Trp-Arg and Tyr-Arg were favourable for drug development. Our comprehensive computational investigation has identified these four dipeptides as promising candidates. These designed and selected dipeptides may further be modified using peptidomimetic and medicinal chemistry tools to develop a novel class of promising antimalarials.

Structural and functional consequences of non-synonymous SNPs within the LAMA2 protein: a molecular dynamics perspective

Clinical phenotypic presentations associated with LAMA2 deficiency have shown a variety of manifestations. LAMA2 mutations are mainly linked to congenital muscular dystrophy, but there is also mounting evidence suggesting their presence in inflammatory breast cancer, laryngopharyngeal squamous cell carcinoma, and ventricular tachycardia related to coronary artery disease and cardiomyopathy. This study examined the structural and functional impacts of 144 non-synonymous single nucleotide polymorphisms (nsSNPs) within the LAMA2 gene. Through multi-tiered sequence and structure-based methods, 11 deleterious and destabilizing mutations were identified (A1362T, E1308Q, E1360G, I1276S, L1195P, M1359T, P1232H, P1238A, P1272L, Y1234H, Y1338C). Further, four mutations (L1195P, Y1234H, P1238A, A1362T), which aligned with conserved positions, were subjected to 500 ns molecular dynamics (MD) simulations. RMSD calculated from MD trajectories highlighted structural disparities between wild-type and mutant forms, with the latter showing greater flexibility. Radius of gyration analysis indicated reduced compactness, solvent accessibility changes suggested unfolding, and hydrogen bond (HB) analysis demonstrated disrupted integrity. The HB analysis revealed disruptions in structural integrity due to diminished hydrogen bonds in mutants. Secondary structure analysis revealed significant alterations in secondary structural content. Principal Component Analysis unveiled increased dynamic behavior in mutants. Gibbs free energy landscape analysis reflected distinct energy minima regions in mutants, indicating structural destabilization. Overall, this study revealed the functional and structural ramifications of nsSNPs in the LAMA2 gene, providing valuable insights into potential disease-causing mutations and warranting future research on understanding LAMA2 associated diseases and disorders.

MIPPIS: protein–protein interaction site prediction network with multi-information fusion

BackgroundThe prediction of protein–protein interaction sites plays a crucial role in biochemical processes. Investigating the interaction between viruses and receptor proteins through biological techniques aids in understanding disease mechanisms and guides the development of corresponding drugs. While various methods have been proposed in the past, they often suffer from drawbacks such as long processing times, high costs, and low accuracy.ResultsAddressing these challenges, we propose a novel protein–protein interaction site prediction network based on multi-information fusion. In our approach, the initial amino acid features are depicted by the position-specific scoring matrix, hidden Markov model, dictionary of protein secondary structure, and one-hot encoding. Simultaneously, we adopt a multi-channel approach to extract deep-level amino acids features from different perspectives. The graph convolutional network channel effectively extracts spatial structural information. The bidirectional long short-term memory channel treats the amino acid sequence as natural language, capturing the protein’s primary structure information. The ProtT5 protein large language model channel outputs a more comprehensive amino acid embedding representation, providing a robust complement to the two aforementioned channels. Finally, the obtained amino acid features are fed into the prediction layer for the final prediction.ConclusionCompared with six protein structure-based methods and six protein sequence-based methods, our model achieves optimal performance across evaluation metrics, including accuracy, precision, F1, Matthews correlation coefficient, and area under the precision recall curve, which demonstrates the superiority of our model.

In Silico Screening and Molecular Dynamics Simulations against Tyrosine-protein Kinase Fyn Reveal Potential Novel Therapeutic Candidates for Bovine Papillomatosis.

Decreased beef productivity due to papillomatosis has led to the development and identification of novel targets and molecules to treat the disease. Protein kinases are promising targets for the design of numerous chemotherapy drugs. This study aimed to screen and design new inhibitors of bovine Fyn, a protein kinase, using structure-based computational methods, such as molecular docking and molecular dynamics simulation (MDS). To carry out the molecular docking analysis, five ligands obtained through structural similarity between active compounds along with the cross-inhibition function between the ChEMBL and Drugbank databases were used. Molecular modeling was performed, and the generated models were validated using PROCHECK and Verify 3D. Molecular docking was performed using Autodock Vina. The complexes formed between Fyn and the three best ligands had their stability assessed by MDS. In these simulations, the complexes were stabilized for 100 ns in relation to a pressure of 1 atm, with an average temperature of 300 k and a potential energy of 1,145,336 kJ/m converged in 997 steps. Docking analyses showed that all selected ligands had a high binding affinity with Fyn and presented hydrogen bonds at important active sites. MDS results support the docking results, as the ligand showed similar and stable interactions with amino acids present at the binding site of the protein. In all simulations, sorafenib obtained the best results of interaction with the bovine Fyn. The results highlight the identification of possible bovine Fyn inhibitors; however, further studies are important to confirm these results experimentally.

A head-to-head comparison of MM/PBSA and MM/GBSA in predicting binding affinities for the CB1 cannabinoid ligands.

The substantial increase in the number of active and inactive-state CB1 receptor experimental structures has provided opportunities for CB1 drug discovery using various structure-based drug design methods, including the popular end-point methods for predicting binding free energies-Molecular Mechanics/Poisson-Boltzmann Surface Area (MM/PBSA) and Molecular Mechanics/Generalized Born Surface Area (MM/GBSA). In this study, we have therefore evaluated the performance of MM/PBSA and MM/GBSA in calculating binding free energies for CB1 receptor. Additionally, with both MM/PBSA and MM/GBSA being known for their highly individualized performance, we have evaluated the effects of various simulation parameters including the use of energy minimized structures, choice of solute dielectric constant, inclusion of entropy, and the effects of the five GB models. Generally, MM/GBSA provided higher correlations than MM/PBSA (rMM/GBSA = 0.433 - 0.652 vs. rMM/PBSA = 0.100 - 0.486) regardless of the simulation parameters, while also offering faster calculations. Improved correlations were observed with the use of molecular dynamics ensembles compared with energyminimized structures and larger solute dielectric constants. Incorporation of entropic terms led to unfavorable results for both MM/PBSA and MM/GBSA for a majority of the dataset, while the evaluation of the various GB models exerted a varying effect on both the datasets. The findings obtained in this study demonstrate the utility of MM/PBSA and MM/GBSA in predicting binding free energies for the CB1 receptor, hence providing a useful benchmark for their applicability in the endocannabinoid system as well as other G protein-coupled receptors. The study utilized the docked dataset (Induced Fit Docking with Glide XP scoring function) from Loo et al., consisting of 46 ligands-23 agonists and 23 antagonists. The equilibrated structures from Loo et al. were subjected to 30ns production simulations using GROMACS 2018 at 300K and 1atm with the velocity rescaling thermostat and the Parinello-Rahman barostat. AMBER ff99SB*-ILDN was used for the proteins, General Amber Force Field (GAFF) was used for the ligands, and Slipids parameters were used for lipids. MM/PBSA and MM/GBSA binding free energies were then calculated using gmx_MMPBSA. The solute dielectric constant was varied between 1, 2, and 4 to study the effect of different solute dielectric constants on the performance of MM/PB(GB)SA. The effect of entropy on MM/PB(GB)SA binding free energies was evaluated using the interaction entropy module implemented in gmx_MMPBSA. Five GB models, GBHCT, GBOBC1, GBOBC2, GBNeck, and GBNeck2, were evaluated to study the effect of the choice of GB models in the performance of MM/GBSA. Pearson correlation coefficients were used to measure the correlation between experimental and predicted binding free energies.

Developing Generalizable Scoring Functions for Molecular Docking: Challenges and Perspectives.

Structure-based drug discovery methods, such as molecular docking and virtual screening, have become invaluable tools in developing novel drugs. At the core of these methods are Scoring Functions (SFs), which predict the binding affinity between ligands and protein targets. This study aims to review and contextualize the challenges and best practices in training novel scoring functions to improve their accuracy and generalizability in predicting protein-ligand binding affinities. Effective training of scoring functions requires careful attention to the quality of training data and methodologies. We emphasize the need for robust training strategies to produce consistent and generalizable SFs. Key considerations include addressing hidden biases and overfitting in machine-learning models, as well as ensuring the use of high-quality, unbiased datasets for both training and evaluation of SFs. Innovative hybrid methods, combining the advantages of empirical and machine-learning approaches, hold promise for outperforming current scoring functions while displaying greater generalizability and versatility.

Alleviating the resolution limit problem in spatial community detection: a local network structure-based method

Spatial community detection plays a crucial role in the analysis of spatially embedded networks. However, most existing methods adopt modularity as an objective function, which may fail to detect small spatial communities (i.e. the well-known resolution-limit problem). To alleviate this problem, a local network structure-based spatially constrained Leiden method was developed. First, the weights of the edges were reset based on the local network structure, which facilitated a clearer delineation between distinct communities. Second, we extended Leiden, an effective community detection method, to spatial community detection content by adding spatial constraints. Experiments on simulated datasets demonstrated that the proposed method is superior to four state-of-the-art methods for detecting spatial communities of different sizes. A case study conducted using the Shenzhen taxi dataset demonstrated that the proposed method outperformed four state-of-the-art methods in revealing urban spatial structures. Notably, the modularity of the spatial communities detected using the proposed method exhibited a marked improvement, nearly doubling that of the four comparative methods in the case study. This study presents a novel and promising framework for detecting spatial communities using modularity.

Structure-Based Training: A Training Method Aimed at Pixel Errors for a Correlation-Coefficient-Based Neural Network.

In research on building a one-shot learning neural network without pre-training using mass data, the limitation on the information obtained from a single training sample downgrades the performance of the network. In order to improve performance and take full advantage of the support set, in this study, we design three kinds of shadow nodes and propose a structure-based training method for a correlation-coefficient-based neural network. This training strategy focuses on branches that are not activated or inactivated as expected. In contrast to existing networks that optimize the parameters using back-propagation, the training method proposed in this paper optimizes the structure of the correlation-coefficient-based network by correcting its pixel errors. For the shadow nodes and training process based on this strategy, the intersection over union (IOU) of a detected target increases by 4.83% in the experiments when using the Fashion-Mnist dataset, increases by 4.02% when using the Omniglot dataset, and increases by 3.89% when using the Cifar-10 dataset. The samples in category "7" wrongly classified as "1" decreased by 27.32% when using the Mnist dataset after training. This training strategy, along with shadow nodes, makes the correlation-coefficient-based network a more practical model and enables the network to develop during the accumulation of reliable samples, thus making it more suitable for simple target detection projects that collect samples over time. Moreover, the shadow nodes and training method proposed in this paper supplement the non-gradient-based parameter-gaining strategy. Additionally, it is a new attempt to explore the imitation of a human's ability to learn a new pattern from a low number of references.

SeqDance: A Protein Language Model for Representing Protein Dynamic Properties.

Proteins perform their functions by folding amino acid sequences into dynamic structural ensembles. Despite the important role of protein dynamics, their complexity and the absence of efficient representation methods have limited their integration into studies on protein function and mutation fitness, especially in deep learning applications. To address this, we present SeqDance, a protein language model designed to learn representation of protein dynamic properties directly from sequence alone. SeqDance is pre-trained on dynamic biophysical properties derived from over 30,400 molecular dynamics trajectories and 28,600 normal mode analyses. Our results show that SeqDance effectively captures local dynamic interactions, co-movement patterns, and global conformational features, even for proteins lacking homologs in the pre-training set. Additionally, we showed that SeqDance enhances the prediction of protein fitness landscapes, disorder-to-order transition binding regions, and phase-separating proteins. By learning dynamic properties from sequence, SeqDance complements conventional evolution- and static structure-based methods, offering new insights into protein behavior and function.

Based on Molecular Docking, Molecular Dynamics Simulation and MM/PB(GB)SA to Study Potential Inhibitorsof PRRSV-Nsp4.

Porcine reproductive and respiratory syndrome (PRRS) is one of the most serious infectious immunosuppressive diseases in the world. The nonstructural protein Nsp4 can be used as an ideal target for anti-PRRSV replication inhibitors. However, little is known about potential inhibitors that target Nsp4 to affect PRRSV replication. The purpose of this study was to screen potential natural inhibitors that affect PRRSV replication by inhibiting Nsp4. Five compounds with strong binding affinity to Nsp4 were selected by structure-based molecular docking method. The complexes of naringin dihydrochalcone (NDC), agathisflavone (AGT), and amentoflavone (AMF) with Nsp4 were stable throughout the molecular dynamics simulation. According to MM/PBSA analysis, the free energies of binding of NDC, AGT, and AMF to Nsp4 were less than-30 Kcal/mol. In conclusion, these three compounds are worthy of further investigation as novel inhibitors of PRRSV. This study provides a theoretical basis for the development of anti-PRRSV natural drugs.

PDNAPred: Interpretable prediction of protein-DNA binding sites based on pre-trained protein language models

Protein-DNA interactions play critical roles in various biological processes and are essential for drug discovery. However, traditional experimental methods are labor-intensive and unable to keep pace with the increasing volume of protein sequences, leading to a substantial number of proteins lacking DNA-binding annotations. Therefore, developing an efficient computational method to identify protein-DNA binding sites is crucial. Unfortunately, most existing computational methods rely on manually selected features or protein structure information, making these methods inapplicable to large-scale prediction tasks. In this study, we introduced PDNAPred, a sequence-based method that combines two pre-trained protein language models with a designed CNN-GRU network to identify DNA-binding sites. Additionally, to tackle the issue of imbalanced dataset samples, we employed focal loss. Our comprehensive experiments demonstrated that PDNAPred significantly improved the accuracy of DNA-binding site prediction, outperforming existing state-of-the-art sequence-based methods. Remarkably, PDNAPred also achieved results comparable to advanced structure-based methods. The designed CNN-GRU network enhances its capability to detect DNA-binding sites accurately. Furthermore, we validated the versatility of PDNAPred by training it on RNA-binding site datasets, showing its potential as a general framework for amino acid binding site prediction. Finally, we conducted model interpretability analysis to elucidate the reasons behind PDNAPred's outstanding performance.

Exploring the potential of structure-based deep learning approaches for T cell receptor design.

Deep learning methods, trained on the increasing set of available protein 3D structures and sequences, have substantially impacted the protein modeling and design field. These advancements have facilitated the creation of novel proteins, or the optimization of existing ones designed for specific functions, such as binding a target protein. Despite the demonstrated potential of such approaches in designing general protein binders, their application in designing immunotherapeutics remains relatively underexplored. A relevant application is the design of T cell receptors (TCRs). Given the crucial role of T cells in mediating immune responses, redirecting these cells to tumor or infected target cells through the engineering of TCRs has shown promising results in treating diseases, especially cancer. However, the computational design of TCR interactions presents challenges for current physics-based methods, particularly due to the unique natural characteristics of these interfaces, such as low affinity and cross-reactivity. For this reason, in this study, we explored the potential of two structure-based deep learning protein design methods, ProteinMPNN and ESM-IF1, in designing fixed-backbone TCRs for binding target antigenic peptides presented by the MHC through different design scenarios. To evaluate TCR designs, we employed a comprehensive set of sequence- and structure-based metrics, highlighting the benefits of these methods in comparison to classical physics-based design methods and identifying deficiencies for improvement.

Combining Data-Driven and Structure-Based Approaches in Designing Dual PARP1-BRD4 Inhibitors for Breast Cancer Treatment.

Poly(ADP-ribose) polymerase 1 (PARP1) inhibitors have revolutionized the treatment of many cancers with DNA-repairing deficiencies via synthetic lethality. Advocated by the polypharmacology concept, recent evidence discovered that a significantly synergistic effect in increasing the death of cancer cells was observed by simultaneously perturbating the enzymatic activities of bromodomain-containing protein 4 (BRD4) and PARP1. Here, we developed a novel cheminformatics approach combined with a structure-based method aiming to facilitate the design of dual PARP1-BRD4 inhibitors. Instead of linking pharmacophores, the developed approach first identified merged pharmacophores (a pool of amide-containing ring systems), from which phenanthridin-6(5H)-one was further prioritized. Based on this starting point, several small molecules were rationally designed, among which HF4 exhibited low micromolar inhibitory activity against BRD4 and PARP1, particularly exhibiting strong inhibition of BRD4 BD1 with an IC50 value of 204 nM. Furthermore, it demonstrated potent antiproliferative effects against breast cancer gene-deficient and proficient breast cancer cell lines by arresting cell cycle progression and impeding DNA damage repair. Collectively, our systematic efforts to design lead-like molecules have the potential to open doors for the exploration of dual PARP1-BRD4 inhibitors as a promising avenue for breast cancer treatment. Furthermore, the developed approach can be extended to systematically design inhibitors targeting PARP1 and other related targets.

Discovering Dually Active Anti-cancer Compounds with a Hybrid AI-structure-based Approach.

Cancer's persistent growth often relies on its ability to maintain telomere length and tolerate the accumulation of DNA damage. This study explores a computational approach to identify compounds that can simultaneously target both G-quadruplex (G4) structures and poly(ADP-ribose) polymerase (PARP)1 enzyme, offering a potential multipronged attack on cancer cells. We employed a hybrid virtual screening (VS) protocol, combining the power of machine learning with traditional structure-based methods. PyRMD, our AI-powered tool, was first used to analyze vast chemical libraries and to identify potential PARP1 inhibitors based on known bioactivity data. Subsequently, a structure-based VS approach selected compounds from these identified inhibitors for their G4 stabilization potential. This two-step process yielded 50 promising candidates, which were then experimentally validated for their ability to inhibit PARP1 and stabilize G4 structures. Ultimately, four lead compounds emerged as promising candidates with the desired dual activity and demonstrated antiproliferative effects against specific cancer cell lines. This study highlights the potential of combining Artificial Intelligence and structure-based methods for the discovery of multitarget anticancer compounds, offering a valuable approach for future drug development efforts.

Lessons learned during the journey of data: from experiment to model for predicting kinase affinity, selectivity, polypharmacology, and resistance.

Recent advances in machine learning (ML) are reshaping drug discovery. Structure-based ML methods use physically-inspired models to predict binding affinities from protein:ligand complexes. These methods promise to enable the integration of data for many related targets, which addresses issues related to data scarcity for single targets and could enable generalizable predictions for a broad range of targets, including mutants. In this work, we report our experiences in building KinoML, a novel framework for ML in target-based small molecule drug discovery with an emphasis on structure-enabled methods. KinoML focuses currently on kinases as the relative structural conservation of this protein superfamily, particularly in the kinase domain, means it is possible to leverage data from the entire superfamily to make structure-informed predictions about binding affinities, selectivities, and drug resistance. Some key lessons learned in building KinoML include: the importance of reproducible data collection and deposition, the harmonization of molecular data and featurization, and the choice of the right data format to ensure reusability and reproducibility of ML models. As a result, KinoML allows users to easily achieve three tasks: accessing and curating molecular data; featurizing this data with representations suitable for ML applications; and running reproducible ML experiments that require access to ligand, protein, and assay information to predict ligand affinity. Despite KinoML focusing on kinases, this framework can be applied to other proteins. The lessons reported here can help guide the development of platforms for structure-enabled ML in other areas of drug discovery.

An artificial intelligence accelerated virtual screening platform for drug discovery

Structure-based virtual screening is a key tool in early drug discovery, with growing interest in the screening of multi-billion chemical compound libraries. However, the success of virtual screening crucially depends on the accuracy of the binding pose and binding affinity predicted by computational docking. Here we develop a highly accurate structure-based virtual screen method, RosettaVS, for predicting docking poses and binding affinities. Our approach outperforms other state-of-the-art methods on a wide range of benchmarks, partially due to our ability to model receptor flexibility. We incorporate this into a new open-source artificial intelligence accelerated virtual screening platform for drug discovery. Using this platform, we screen multi-billion compound libraries against two unrelated targets, a ubiquitin ligase target KLHDC2 and the human voltage-gated sodium channel NaV1.7. For both targets, we discover hit compounds, including seven hits (14% hit rate) to KLHDC2 and four hits (44% hit rate) to NaV1.7, all with single digit micromolar binding affinities. Screening in both cases is completed in less than seven days. Finally, a high resolution X-ray crystallographic structure validates the predicted docking pose for the KLHDC2 ligand complex, demonstrating the effectiveness of our method in lead discovery.

Ligand and structure-based virtual screening approaches in drug discovery: minireview.

The compilation of ligand and structure-based molecular modeling methods has become an important practice in virtual screening applied to drug discovery. This systematic review addresses and ranks various virtual screening strategies to drive the selection of the optimal method for studies that have as their starting point a multi-ligand investigation and investigation based on the protein structure of a therapeutic target. This study shows examples of applications and an evaluation based on the objective and problematic of a series of virtual screening studies present in the ScienceDirect® database. The results showed that the molecular docking technique is widely used in scientific production, indicating that approaches that use protein structure as a starting point are the most promising strategy for drug discovery that relies on virtual screening-based research.

Machine Learning Guided AQFEP: A Fast and Efficient Absolute Free Energy Perturbation Solution for Virtual Screening.

Structure-based methods in drug discovery have become an integral part of the modern drug discovery process. The power of virtual screening lies in its ability to rapidly and cost-effectively explore enormous chemical spaces to select promising ligands for further experimental investigation. Relative free energy perturbation (RFEP) and similar methods are the gold standard for binding affinity prediction in drug discovery hit-to-lead and lead optimization phases, but have high computational cost and the requirement of a structural analog with a known activity. Without a reference molecule requirement, absolute FEP (AFEP) has, in theory, better accuracy for hit ID, but in practice, the slow throughput is not compatible with VS, where fast docking and unreliable scoring functions are still the standard. Here, we present an integrated workflow to virtually screen large and diverse chemical libraries efficiently, combining active learning with a physics-based scoring function based on a fast absolute free energy perturbation method. We validated the performance of the approach in the ranking of structurally related ligands, virtual screening hit rate enrichment, and active learning chemical space exploration; disclosing the largest reported collection of free energy simulations to date.

A Multibranch Neural Network for Drug-Target Affinity Prediction Using Similarity Information.

Predicting drug-target affinity (DTA) is beneficial for accelerating drug discovery. In recent years, graph structure-based deep learning models have garnered significant attention in this field. However, these models typically handle drug or target protein in isolation and only extract the molecular structure information on the drug or protein itself. To address this limitation, existing network-based models represent drug-target interactions or affinities as a knowledge graph to capture the interaction information. In this study, we propose a novel solution. Specifically, we introduce drug similarity information and protein similarity information into the field of DTA prediction. Moreover, we propose a network framework that autonomously extracts similarity information, avoiding reliance on knowledge graphs. Based on this framework, we design a multibranch neural network called GASI-DTA. This network integrates similarity information, sequence information, and molecular structure information. Comprehensive experimental results conducted on two benchmark data sets and three cold-start scenarios demonstrate that our model outperforms state-of-the-art graph structure-based methods in nearly all metrics. Furthermore, it exhibits significant advantages over existing network-based models, outperforming the best of them in the majority of metrics. Our study's code and data are openly accessible at http://github.com/XiaoLin-Yang-S/GASI-DTA.

Virtual Screening and Molecular Docking: Discovering Novel METTL3 Inhibitors.

Methyltransferase-like 3 (METTL3) is an RNA methyltransferase that catalyzes the N6 -methyladenosine (m6A) modification of mRNA in eukaryotic cells. Past studies have shown that METTL3 is highly expressed in various cancers and is closely related to tumor development. Therefore, METTL3 inhibitors have received widespread attention as effective treatments for different types of tumors. This study proposes a hybrid high-throughput virtual screening (HTVS) protocol that combines structure-based methods with geometric deep learning-based DeepDock algorithms. We identified unique skeleton inhibitors of METTL3 from our self-built internal database. Among them, compound C3 showed significant inhibitory activity on METTL3, and further molecular dynamics simulations were performed to provide more details about the binding conformation. Overall, our research demonstrates the effectiveness of hybrid virtual algorithms, which is of great significance for understanding the biological functions of METTL3 and developing treatment methods for related diseases.