Molecular glues stabilize transient protein-protein interactions by creating a new binding interface, often turning a non-functional complex into a targeted degradation signal. Our service models these glue-induced interfaces from the ground up — from ternary complex prediction and glue efficiency scoring to interface hotspot mapping and rational glue optimization. We combine large-scale docking, molecular dynamics, and free energy analysis with targeted in vitro validation to de-risk glue discovery.
We use ensemble docking and template-based modeling to predict the 3D structure of the ternary complex (target–glue–E3 ligase or partner protein), focusing on the neo-interface created by the glue molecule.
We compute the binding free energy of the ternary complex and decompose it into protein–protein and protein–glue contributions, quantifying cooperativity (α) and predicting glue efficiency.
Explicit-solvent MD simulations capture the dynamic stability of the glue-induced interface, revealing transient contacts and water-mediated interactions that static structures miss.
We suggest chemical modifications to the glue molecule or mutations on the protein partners to enhance ternary complex stability, specificity, or degradation efficiency.
For targeted protein degradation applications, we correlate glue-induced interface metrics with predicted degradation efficiency, leveraging published ternary complex features.
We design focused wet-lab experiments — SPR, ITC, or cellular degradation assays — to validate our top computational hits and feed data back into the model.
We use molecular docking and induced-fit protocols to generate and refine glue poses within the binary complexes, then build ternary complexes using interface-driven docking with in-house algorithms and established protein-protein docking approaches.
We apply end-point free energy methods to estimate binding cooperativity and, for critical decisions, use free energy perturbation (FEP) to calculate relative binding affinities of glue analogs.
We run 100–500 ns MD simulations on top ternary complex poses to assess interface stability, identify dynamic hotspots, and detect glue-induced conformational changes.
We leverage published ternary complex structures and degradation data to train ML models that predict glue efficiency from interface descriptors, helping prioritize novel glue scaffolds.
We model residue interaction networks to identify communication pathways between the glue binding site and distal functional regions, especially relevant for E3 ligase activation.
Where available, we integrate proteomics, CRISPR screening, or transcriptomics data to contextualize glue-induced interface modeling within a broader cellular response.
We review your target protein, E3 ligase (or partner), and available glue warheads. If no structure is available, we build homology models or use AlphaFold2 predictions to prepare the binding-competent states.
We dock the glue molecule into the target protein or E3 ligase, using ensemble docking to account for receptor flexibility. Top poses are refined with induced-fit protocols.
We construct ternary complexes by docking the second protein partner onto the glue-bound complex. Thousands of poses are scored by interface complementarity, buried surface area, and glue-mediated contacts.
Top ternary complex poses are subjected to explicit-solvent MD simulation (100–500 ns) followed by MM/PBSA or FEP calculations to quantify interface stability and cooperativity.
We map energetic hotspots at the glue-induced interface, identify residues critical for complex stability, and suggest glue modifications or protein mutations to improve cooperativity or specificity.
We recommend focused SPR, ITC, or cellular assays to validate top predictions. Experimental results are fed back to refine our computational models, creating a closed-loop design cycle.
| Challenge | Our Modeling Strategy | Key Outputs | Impact |
|---|---|---|---|
| I have a glue warhead but no ternary complex structure. | Ensemble docking + MD refinement + interface scoring | Predicted ternary complex models, glue efficiency scores | Prioritize glue candidates for synthesis |
| I need to improve cooperativity between target and E3 ligase. | Per-residue energy decomposition + glue modification prediction | Hotspot maps, suggested glue analogs with predicted ΔΔG | Design more potent molecular glues |
| I want to understand why my glue fails to degrade a target. | Ternary complex geometry + lysine proximity + MD stability | Ubiquitination feasibility score, interface destabilization sites | Guide linker or glue optimization |
| I need to design mutations to validate the induced interface. | Alanine scanning + free energy perturbation | Predicted ΔΔG for interface mutations | Design clean wet-lab controls |
| I want to identify novel glue scaffolds from a compound library. | Virtual screening + ML-based glue efficiency prediction | Ranked list of glue-like compounds with predicted ternary stability | Expand chemical space for glue discovery |
Model glue-induced interfaces between a target protein and an E3 ligase to predict degradation efficiency and guide glue optimization.
Design molecular glues that stabilize or create new signaling complexes, e.g., 14-3-3 protein interactions or immune receptor clustering.
Identify small molecules that can reinforce weakened protein-protein interfaces caused by disease mutations, acting as molecular glues for interface repair.
Develop molecular glues as chemical tools to acutely control protein complex formation in cells, enabling target validation studies.
Model glues that induce or stabilize immune receptor complexes, potentially enhancing T-cell activation or checkpoint blockade.
Explore glue-induced proximity between a target protein and a functional effector for novel therapeutic modalities.
Molecular glue discovery is notoriously challenging — the induced interfaces are often flat, dynamic, and difficult to predict. Our service combines state-of-the-art computational modeling with a pragmatic wet-lab validation loop: we use SPR, ITC, and cellular assays to test our top predictions, then feed the data back into our models. This closed-loop approach increases the success rate of finding potent and selective molecular glues.
Objective: identify molecular glues that induce ternary complex formation between a target protein and a poorly characterized E3 ligase.
Objective: improve cooperativity of an existing glue by 10-fold through rational chemical modification.
Objective: explain how clinical mutations in the target protein disrupt glue-induced degradation.
PROTACs are bifunctional molecules with two distinct binding warheads connected by a linker — one for the target protein and one for the E3 ligase. Molecular glues are typically smaller, monovalent molecules that induce or stabilize a protein-protein interface by binding to one protein and creating a new interaction surface for the partner. Glues are often serendipitously discovered, but computational modeling can guide their rational design.
Yes. We use a combination of homology modeling, AlphaFold2 predictions, and ensemble docking to generate ternary complex models. While the accuracy depends on the quality of the input structures and the glue's binding mode, our MD refinement and free energy analysis help prioritize the most plausible models for experimental testing.
We analyze several features of the modeled ternary complex: the distance between the target protein's lysine residues and the E3 ligase's ubiquitination machinery, the stability of the interface (from MD and MM/PBSA), and the geometry of the complex. These features are combined into a degradation feasibility score, which we calibrate against published degradation data where available.
We typically recommend SPR or ITC to measure ternary complex formation and cooperativity, followed by cellular degradation assays (e.g., Western blot) if the goal is targeted protein degradation. For interface validation, we suggest alanine scanning or site-directed mutagenesis of predicted hotspot residues, followed by binding or degradation assays.
A focused project — including ternary complex prediction, MD simulation, and hotspot analysis — typically takes 3–5 weeks. If extensive virtual screening or FEP calculations are required, the timeline may extend to 6–8 weeks. We provide a detailed timeline during project scoping.
Yes. We can perform virtual screening of glue-like compound libraries against the binary complex, followed by ternary complex assembly and scoring. We also offer machine learning-based glue efficiency prediction to prioritize novel scaffolds.
Share your target protein, E3 ligase, and glue warhead — or just a hypothesis — and our team will design a tailored molecular glue modeling plan. We combine rigorous computational chemistry with wet-lab validation to help you discover and optimize molecular glues for targeted protein degradation and beyond.
Submit Your Project DetailsSubmit your project details below, and our team will respond within 24 hours.
Talk to our technical team about your project!
I Want To Talk
