Issue #85: BA-Pred and RMSD-Pred: Integrated Graph Neural Network Models for Accurate Protein-Ligand Binding Affinity and Binding Pose Prediction.

Signal of the Day

BA-Pred and RMSD-Pred: Integrated Graph Neural Network Models for Accurate Protein-Ligand Binding Affinity and Binding Pose Prediction.

Accurate prediction of protein-ligand bound poses and their affinities is essential in structure-based drug discovery. Here, we present an integrated deep-learning framework that disentangles the two core tasks─affinity estimation and pose evaluation─within complementary graph neural network architectures. BA-Pred is for predicting binding affinity and RMSD-Pred is for binding pose assessment, predicting the root-mean-squared deviation of ligand poses from crystal structures. Both models employ a Gated Graph Convolutional Network with Learnable Structural Positional Encoding (GatedGCN-LSPE) architecture to capture spatial and chemical dependencies across protein-ligand graphs. BA-Pred achieved state-of-the-art scoring power on the CASF-2016 benchmark with a root-mean-squared error of 1.10 p K d , while RMSD-Pred exhibited strong docking power with a top-1 success rate of 96%, comparable to the best reported deep-learning scoring functions. The robust generalization capability of RMSD-Pred was further validated on the external Astex diverse set and PoseBusters benchmarks, where it significantly improved the pose selection success rates of AutoDock-GPU by up to 33.1%. The accuracy of our methodology was demonstrated on pharmaceutical targets in the 16th Critical Assessment of Structure Prediction, where our approach ranked second in the ligand binding affinity prediction category. By using our models, an integrated pipeline was developed for virtual screening, where pose selection was performed with RMSD-Pred and binding affinities were predicted with BA-Pred. This combined approach demonstrated robust screening performance, achieving an enrichment factor (EF) 1% of 21.1 on the CASF-2016 benchmark. Furthermore, on the LIT-PCBA benchmark, rescoring poses docked by AutoDock-GPU with our pipeline significantly improved the EF 1% from 2.18 to 3.19. These various benchmark results demonstrate that our graph-neural network models show good and balanced performance in diverse protein-ligand interaction prediction tasks. Thus, we expect that our models will serve as a promising framework to accelerate the drug discovery process.

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

Also Worth Reading

Enhancing CYP450-Ligand Binding Predictions: A Comparative Analysis of Ligand-Based and Hybrid Machine Learning Models.

Predicting cytochrome P450 (CYP450) ligand binding is critical in early-stage drug discovery as CYP450-mediated metabolism profoundly influences drug efficacy, safety, and adverse reaction risks. However, experimental determination of CYP450-ligand interactions remains resource- and time-intensive, underscoring the need for robust computational alternatives. While ligand-based methods are commonly employed, they often fail to fully account for structural intricacies governing protein-ligand interactions. To address this gap, we developed a hybrid machine learning framework integrating ligand descriptors, protein descriptors, and protein-ligand interaction descriptors that include molecular docking-derived parameters, rescoring function components from multiple algorithms, and structural interaction fingerprints (SIFt). Evaluated on CYP1A2 and CYP17A1 isoforms, our model demonstrated superior predictive accuracy in cross-validation compared with stand-alone molecular docking and ligand-based approaches. Furthermore, benchmarking against state-of-the-art tools (SwissADME and ADMETlab 3.0) revealed enhanced performance in binding prediction. This work establishes a versatile framework for advancing computational tools to prioritize CYP450 binding assessments during drug discovery.

Evaluating zero-shot prediction of monomeric protein design success by AlphaFold, ESMFold, and ProteinMPNN.

De novo protein design has enabled the creation of proteins with diverse functionalities that are not found in nature. Despite recent advances, experimental success rates remain inconsistent and context-dependent, posing a bottleneck for broader applications of de novo design. To overcome this, structure and sequence prediction models have been applied to assess design quality prior to experimental testing to save time and resources. In this study, we examined the extent to which AlphaFold, Protein MPNN, and ESMFold can discriminate between experimentally successful and unsuccessful designs. We first curated a benchmark dataset of 614 experimentally characterized de novo designed monomers from 11 different design studies between 2012 and 2021. All predictive models demonstrated moderate ability to discriminate experimental successes (expressed, soluble, monomeric, and fold with the correct secondary structure) from failures. Still, many failed designs have better confidence metrics than successful designs, and confidence metrics were topology-dependent. Among all computational models evaluated, ESMFold average predicted local-distance difference test (pLDDT) yielded the best individual performance at distinguishing between successful and unsuccessful designs. A logistic regression model combining all confidence metrics provided only modest improvement over ESMFold pLDDT alone. Overall, these results show that these models can serve as an initial filtering strategy prior to experimental validation; however, their utility at accurately predicting experimentally successful designs remains limited without task-specific training.

Integrated network pharmacology and AlphaFold modeling reveal ESR1 as a key target of Huanglian Jiedu decoction for ameliorating sepsis-induced coagulopathy.

Ethnopharmacological relevance Huanglian Jiedu Decoction (HLJDD) was initially documented in the Elbow Reserve Emergency Prescription, an ancient Chinese medical text created in the Eastern Jin Dynasty. This decoction is composed of four herbs: Huanglian (HL, Coptis chinensis Franch.), Zhizi (ZZ, Gardenia jasminoides Ellis), Huangqin (HQ, Scutellaria baicalensis Georgi), and Huangbo (HB, Phellodendron chinense Schneid.). Modern pharmacological research shows that HLJDD exerts multiple therapeutic effects, including antibacterial and anti-inflammatory properties, improvement of coagulation function, and amelioration of cerebral ischemia. Aims of the study Sepsis-induced coagulopathy (SIC) is a critical complication associated with sepsis, characterized by disruptive coagulation. This study evaluated the efficacy of HLJDD in alleviating SIC and elucidated its potential mechanisms through network pharmacology, molecular dynamics, and AlphaFold. Materials and methods The cecal ligation and puncture (CLP) method was used to establish a septic rat model. HLJDD was administered as an intervention. Mortality rates, vital signs, histopathological testing of the lung and kidney, blood cell counts, and coagulation function were assessed to evaluate the severity of coagulation disorders and inflammatory injury across the groups. Network pharmacology was used to identify candidate targets, and the potential mechanism of HLJDD in alleviating SIC was verified by ELISA, RT-qPCR and Western blot. AlphaFold 3 and molecular dynamics simulations were used to predict the potential regulatory mechanisms of ESR1 by the major compound of HLJDD. Results This study demonstrated that HLJDD improved coagulation dysfunction, reduced inflammatory response and lung and kidney pathological damage in CLP-induced sepsis rats, and significantly improved survival rate. The major compounds of HLJDD were identified by integrated network pharmacology and UPLC-Q-TOF-MS. ESR1, IL-6, and CXCL8 were identified as potential targets for HLJDD in alleviating SIC. In the validation of the predicted results, HLJDD restored reduced ESR1 expression in the lungs and kidneys of sepsis rats and exhibited good regulatory effects on IL-6 and CINC-1 (a functional CXCL8 analog in rats). Furthermore, quercetin was identified as the major compound of HLJDD. Molecular docking and molecular dynamics simulations suggested that quercetin has a good binding affinity for ESR1. Based on AlphaFold 3 structural modeling analysis, quercetin may alleviate SIC by mediating the regulation of ESR1 recognition of estrogen response elements (EREs). Conclusion In summary, HLJDD alleviates SIC by reducing PT, APTT levels and increasing the FIB contents. Our study indicates that the mechanism involves the upregulation of ESR1, which enhances ESR1-ERE binding to suppress inflammation and microthrombosis. This elucidates a specific mode of action for HLJDD, highlighting its potential value in SIC treatment.

Research & AI Updates

• AI Transforms Drug Discovery And Development - Let’s Data Science — AI Transforms Drug Discovery And Development    Let’s Data Science.
• The ‘2 AM question’ that Elon Musk and Demis Hassabis can’t stop arguing about - MSN — The ‘2 AM question’ that Elon Musk and Demis Hassabis can’t stop arguing about    MSN.

From the Industry

• Trinity Biotech and University at Buffalo Partner to Commercialize Sjögren’s Syndrome Biomarkers - Clinical Lab Products — Trinity Biotech and University at Buffalo Partner to Commercialize Sjögren’s Syndrome Biomarkers    Clinical Lab Products.
• South Korea streamlines biosimilar review, clinical trials - BioWorld MedTech — South Korea streamlines biosimilar review, clinical trials    BioWorld MedTech.
• Evommune bucked biotech IPO drought and is focused on the long game - Pharma Voice — Evommune bucked biotech IPO drought and is focused on the long game    Pharma Voice.
• Therapeutic Alternatives to Recombinant Biologics: Mechanistic Framework, Clinical Evidence, and Selection Guidance - Dove Medical Press — Therapeutic Alternatives to Recombinant Biologics: Mechanistic Framework, Clinical Evidence, and Selection Guidance    Dove Medical Press.
• Takeda tears up Denali partnership, returning dementia asset amid restructuring - fiercebiotech.com — Takeda tears up Denali partnership, returning dementia asset amid restructuring    fiercebiotech.com.
• Partnership Joins National Symposium On America’s Biotech Moment - Greater Houston Partnership — Partnership Joins National Symposium On America’s Biotech Moment    Greater Houston Partnership.
• March M&A surge triggers high expectations for 2026 - Fierce Pharma — March M&A surge triggers high expectations for 2026    Fierce Pharma.

Quick Reads

Molecular docking, MM/GBSA, FEP/MD, and DFT/MM MD studies on predicting binding affinity of carbonic anhydrase II inhibitors.

Human carbonic anhydrase II (hCAII) is one of the zinc-containing metalloenzymes that catalyzes various hydration reactions. Read more →

Atomic resolution ensembles of intrinsically disordered proteins with Alphafold.

Intrinsically disordered proteins are ubiquitous in biological systems and play essential roles in a wide range of biological processes and diseases. Read more →

Tongnao Decoction Exerts a Treatment Effect on Ischemic Stroke by IL-6/PI3K/Akt/GSK-3β Pathway: Based on the Network Pharmacology and Molecular Docking.

Ischemic stroke (IS) is an important disease leading to high disability and mortality, and the current clinical treatment is limited. Read more →

Towards protein folding pathways by reconstructing protein residue networks with a policy-driven model

A method that reconstructs protein residue networks using suitable node selection and edge recovery policies produced numerical observations that correlate strongly (Pearson’s correlation coefficient < -0.83) with published folding rates for 52 two-state folders and 21 multi-state folders; correlations are also strong at the fold-family level. Read more →

Molecular Insights into Identification of Natural AKT1/mTOR Signaling Inhibitors from Veratrum Viride-Derived Alkaloids for Breast Cancer Treatment: A Comprehensive Analysis Using Network Pharmacology, Molecular Docking, and Molecular Dynamics.

Breast cancer (BC) is a complex illness that affects millions of women globally. Read more →

Multiscale anisotropic scaffolds enable a biomimetic electro-mechanical myocardial platform for drug discovery and heart repair.

The core challenge in engineering functional cardiac tissue in vitro is the lack of an integrated platform that simultaneously provides multiscale anisotropic topography and conductive signaling, leading to poor cellular alignment, weak electromechanical coupling, and asynchronous contraction. Read more →

Curcumin-loaded beta-glucan nanoparticles: preparation, physicochemical properties, and molecular modeling.

This study introduces the application of wet milling, a well-established top-down technique, as a novel approach for preparing beta-glucan nanoparticles (BG-NPs) as carriers for poorly soluble drugs, using curcumin (CC) as a model compound. Read more →

Is scaffold hopping possible in machine learning using the electronic-structure-informatics (ESI) descriptor set? an application to natural-product-based drug discovery of α-glucosidase inhibitors.

The electronic-structure informatics (ESI) descriptor set was applied to discover novel α-glucosidase inhibitors from a natural product (NP) database. Read more →

Pipeline Tip

Use GPU-accelerated MD refinement to lift model quality in under 2 hours.

Resources & Tools

• Dataset: MGnify - Metagenomics resource for microbiome sequence data.
• Dataset: PDBbind - Binding affinity data with 3D structures of protein-ligand complexes.
• Tool: OmegaFold - Structure prediction from single sequences with rapid inference. View all tools →
• Tool: Foldseek - Ultra-fast structural search and clustering engine. View all tools →
• Event: Protein Design Hub (LinkedIn Group) (Ongoing)
• Event: Structural Biology Events (Open)
• Job: Bioinformatics Software Engineer Jobs, Employment - Indeed at Indeed Jobs
• Job: 1,000 Bioinformatics Research Job Vacancies in Mahadevapura, Bengaluru, Karnataka - Indeed at Indeed Jobs

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