Explore the vast range of titles available.

Accurate prediction of the binding affinity of a protein-ligand complex is essential for efficient and successful rational drug design. Therefore, many binding affinity prediction methods have been developed. In recent years, since deep learning technology has become powerful, it is also implemented to predict affinity. In this work, a new neural network model that predicts the binding affinity of a protein-ligand complex structure is developed. Our model predicts the binding affinity of a complex using the ensemble of multiple independently trained networks that consist of multiple channels of 3-D convolutional neural network layers. Our model was trained using the 3772 protein-ligand complexes from the refined set of the PDBbind-2016 database and tested using the core set of 285 complexes. The benchmark results show that the Pearson correlation coefficient between the predicted binding affinities by our model and the experimental data is 0.827, which is higher than the state-of-the-art binding affinity prediction scoring functions. Additionally, our method ranks the relative binding affinities of possible multiple binders of a protein quite accurately, comparable to the other scoring functions. Last, we measured which structural information is critical for predicting binding affinity and found that the complementarity between the protein and ligand is most important.

Proteins interact with small molecules to modulate several important cellular functions. Many acute diseases were cured by small molecule binding in the active site of protein either by inhibition or activation. Currently, there are several docking programs to estimate the binding position and the binding orientation of protein–ligand complex. Many scoring functions were developed to estimate the binding strength and predict the effective protein–ligand binding. While the accuracy of current scoring function is limited by several aspects, the solvent effect, entropy effect, and multibody effect are largely ignored in traditional machine learning methods. In this paper, we proposed a new deep neural network-based model named DeepBindRG to predict the binding affinity of protein–ligand complex, which learns all the effects, binding mode, and specificity implicitly by learning protein–ligand interface contact information from a large protein–ligand dataset. During the initial data processing step, the critical interface information was preserved to make sure the input is suitable for the proposed deep learning model. While validating our model on three independent datasets, DeepBindRG achieves root mean squared error (RMSE) value of pKa (−logK d or −logK i ) about 1.6–1.8 and R value around 0.5–0.6, which is better than the autodock vina whose RMSE value is about 2.2–2.4 and R value is 0.42–0.57. We also explored the detailed reasons for the performance of DeepBindRG, especially for several failed cases by vina. Furthermore, DeepBindRG performed better for four challenging datasets from DUD.E database with no experimental protein–ligand complexes. The better performance of DeepBindRG than autodock vina in predicting protein–ligand binding affinity indicates that deep learning approach can greatly help with the drug discovery process. We also compare the performance of DeepBindRG with a 4D based deep learning method “pafnucy”, the advantage and limitation of both methods have provided clues for improving the deep learning based protein–ligand prediction model in the future.

Accurately estimating the binding strength between proteins and ligands is fundamental in the field of pharmaceutical research and innovation. Previous research has largely concentrated on 1D or 2D molecular descriptors, often neglecting the pivotal 3D features of molecules that profoundly impact drug properties and target binding. This oversight has resulted in diminished predictive performance in molecule-related analyses. A comprehensive grasp of molecular properties necessitates the integration of both local and global molecular information. In this paper, we introduce a deep-learning model, termed PLGAs, which represents molecular systems as graphs based on the three-dimensional configurations of protein–ligand complexes. PLGAs consist of two components: Graph Convolution Networks (GCN) and a Global Attention Mechanism (GAM) network. Specifically, GCNs learn both the graph structure and node attribute information, capturing local and global information to better represent node features. GAM is then used to gather interactive edges by reducing information loss and amplifying global interactions. PLGAs were tested on the standard PDBbind refined set (v.2019) and core set (v.2016). The model demonstrated a Spearman’s correlation coefficient of 0.823 on the refined set and an RMSE (Root Mean Square Error) of 1.211 kcal/mol between experimental and predicted affinities on the core set, surpassing several advanced contemporary binding affinity prediction methods. We further evaluated the efficacy of various components within our model, and the marked improvements in accuracy underscore the potential of PLGAs to significantly enhance the drug development process. Python scripts implementing various components of models are available at https://github.com/ligaili01/PLGAs.

Serine/threonine protein kinases (CK2, PIM-1, RIO1) are constitutively active, highly conserved, pleiotropic, and multifunctional kinases, which control several signaling pathways and regulate many cellular functions, such as cell activity, survival, proliferation, and apoptosis. Over the past decades, they have gained increasing attention as potential therapeutic targets, ranging from various cancers and neurological, inflammation, and autoimmune disorders to viral diseases, including COVID-19. Despite the accumulation of a vast amount of experimental data, there is still no “recipe” that would facilitate the search for new effective kinase inhibitors. The aim of our study was to develop an effective screening method that would be useful for this purpose. A combination of Density Functional Theory calculations and molecular docking, supplemented with newly developed quantitative methods for the comparison of the binding modes, provided deep insight into the set of desirable properties responsible for their inhibition. The mathematical metrics helped assess the distance between the binding modes, while heatmaps revealed the locations in the ligand that should be modified according to binding site requirements. The Structure-Binding Affinity Index and Structural-Binding Affinity Landscape proposed in this paper helped to measure the extent to which binding affinity is gained or lost in response to a relatively small change in the ligand’s structure. The combination of the physico-chemical profile with the aforementioned factors enabled the identification of both “dead” and “promising” search directions. Tests carried out on experimental data have validated and demonstrated the high efficiency of the proposed innovative approach. Our method for quantifying differences between the ligands and their binding capabilities holds promise for guiding future research on new anti-cancer agents.

Recently, we identified a novel mechanism of enzyme inhibition in N-myristoyltransferases (NMTs), which we have named ‘inhibitor trapping’. Inhibitor trapping occurs when the protein captures the small molecule within its structural confines, thereby preventing its free dissociation and resulting in a dramatic increase in inhibitor affinity and potency. Here, we demonstrate that inhibitor trapping also occurs in the kinases. Remarkably, the drug imatinib, which has revolutionized targeted cancer therapy, is entrapped in the structure of the Abl kinase. This effect is also observed in p38α kinase, where inhibitor trapping was found to depend on a ‘magic’ methyl group, which stabilizes the protein conformation and increases the affinity of the compound dramatically. Altogether, these results suggest that inhibitor trapping is not exclusive to N-myristoyltransferases, as it also occurs in the kinase family. Inhibitor trapping could enhance the binding affinity of an inhibitor by thousands of times and is as a key mechanism that plays a critical role in determining drug affinity and potency.

A methylene blue (MB) indicator embedded in sodium alginate (SA) film was previously examined for detecting active oxygen species. In a previous study, spectrometry was used to identify and characterize the MB/SA complex. However, the decolorization mechanism was not fully assessed. In this study, our aim is to conduct computational calculations at the B3LYP/6-31G(d) level to clarify the exact types and positions of the interaction that cause the decolorization in MB. The results demonstrate that MB/SA interacts with carboxylates (-COO(superscript)-(superscript)) of SA and the N, C, and S atoms of MB, confirming previous experimental observations.

The assessment of binding affinity between ligands and the target proteins plays an essential role in drug discovery and design process. As an alternative to widely used scoring approaches, machine learning methods have also been proposed for fast prediction of the binding affinity with promising results, but most of them were developed as all-purpose models despite of the specific functions of different protein families, since proteins from different function families always have different structures and physicochemical features. In this study, we proposed a random forest method to predict the protein–ligand binding affinity based on a comprehensive feature set covering protein sequence, binding pocket, ligand structure and intermolecular interaction. Feature processing and compression was respectively implemented for different protein family datasets, which indicates that different features contribute to different models, so individual representation for each protein family is necessary. Three family-specific models were constructed for three important protein target families of HIV-1 protease, trypsin and carbonic anhydrase respectively. As a comparison, two generic models including diverse protein families were also built. The evaluation results show that models on family-specific datasets have the superior performance to those on the generic datasets and the Pearson and Spearman correlation coefficients ( R p and Rs ) on the test sets are 0.740, 0.874, 0.735 and 0.697, 0.853, 0.723 for HIV-1 protease, trypsin and carbonic anhydrase respectively. Comparisons with the other methods further demonstrate that individual representation and model construction for each protein family is a more reasonable way in predicting the affinity of one particular protein family.

LIN28A is a conserved RNA‐binding protein that inhibits the biogenesis of let‐7 microRNAs, thus promoting cancer progression. However, mechanisms underlying the activation of the LIN28A‐let‐7 signaling pathway remain poorly understood. Here, we show that LIN28A is SUMOylated in vivo and in vitro at K15, which is increased by hypoxia but reduced by chemotherapy drugs such as Cisplatin and Paclitaxel. SUMOylation of LIN28A aggravates its inhibition of let‐7 maturation, resulting in a stark reduction in let‐7, which promotes cancer cell proliferation, migration, invasion, and tumor growth in vivo. Mechanistically, SUMOylation of LIN28A increases its binding affinity with the precursor let‐7 (pre‐let‐7), which subsequently enhances LIN28A‐mediated recruitment of terminal uridylyltransferase TUT4 and simultaneously blocks DICER processing of pre‐let‐7, thereby reducing mature let‐7 production. These effects are abolished in SUMOylation‐deficient mutant LIN28A‐K15R. In summary, these findings shed light on a novel mechanism by which SUMOylation could regulate the LIN28A‐let‐7 pathway in response to cellular stress in cancer cells. This study demonstrated that SUMOylation of LIN28A increases the binding affinity of LIN28A with pre‐let‐7, subsequently to promote TUT4‐mediated uridylation and block DICER processing of pre‐let‐7, thus leading to the reduction of mature let‐7. SUMOylation of Lin28A is regulated in response to cancer cellular stresses such as hypoxia and chemotherapy drug treatment, which have implications in cancer prognosis and therapy.

Accurate prediction of drug‐target interactions constitutes a crucial foundation for drug discovery. DualPG‐DTA is presented, a general framework for binding affinity prediction that integrates two pre‐trained language models to generate atomic‐level molecular representations and residue‐level protein embeddings. The architecture constructs dual molecular‐protein graphs processed through dedicated graph neural networks equipped with dynamic attention mechanisms to extract context‐aware sequence‐level features, which are fused via a multimodal module for affinity predictions. Benchmark results show that DualPG‐DTA consistently outperforms existing models across all metrics. Applied to CDK9 inhibitor discovery, the framework is used to develop robust regression/classification models and identified compound C1 as a novel CDK9 inhibitor with an IC 50 of 1.2 nM. C1 demonstrates exceptional CDK family selectivity alongside optimal pharmacokinetic properties, including prolonged half‐life, adequate clearance, robust plasma exposure, and oral bioavailability. Notably, oral C1 demonstrated potent antitumor efficacy in a Venetoclax‐resistant MV4‐11 acute myeloid leukemia (AML) xenograft model, with concurrent demonstration of favorable tolerability and safety profiles. Collectively, the study not only establishes a unified framework for precise binding affinity prediction but also identifies C1 as a highly promising therapeutic lead targeting CDK9 to conquer Venetoclax resistance in AML.

The epidermal growth factor receptor 2, ERBB2, is a well-validated target for cancer diagnostics and therapy. Recent studies suggest that the over-expression of this receptor in various cancers might also be exploited for antibody-based payload delivery, e.g. antibody drug conjugates. In such strategies, the full-length antibody format is probably not required for therapeutic effect and smaller tumor-specific affinity proteins might be an alternative. However, small proteins and peptides generally suffer from fast excretion through the kidneys, and thereby require frequent administration in order to maintain a therapeutic concentration. In an attempt aimed at combining ERBB2-targeting with antibody-like pharmacokinetic properties in a small protein format, we have engineered bispecific ERBB2-binding proteins that are based on a small albumin-binding domain. Phage display selection against ERBB2 was used for identification of a lead candidate, followed by affinity maturation using second-generation libraries. Cell surface display and flow-cytometric sorting allowed stringent selection of top candidates from pools pre-enriched by phage display. Several affinity-matured molecules were shown to bind human ERBB2 with sub-nanomolar affinity while retaining the interaction with human serum albumin. Moreover, parallel selections against ERBB2 in the presence of human serum albumin identified several amino acid substitutions that dramatically modulate the albumin affinity, which could provide a convenient means to control the pharmacokinetics. The new affinity proteins competed for ERBB2-binding with the monoclonal antibody trastuzumab and recognized the native receptor on a human cancer cell line. Hence, high affinity tumor targeting and tunable albumin binding were combined in one small adaptable protein.