Ref A5
Computational Service
Confidence Reported
Scoped After Feasibility Review
Service Overview
Even with deep-learning prediction available, template-based modelling remains the better route in several situations: when a close experimental template exists in the exact conformational state you care about, when a specific ligand-bound form must be reproduced, or when a predicted model needs its loops, termini and side chains rebuilt to a usable standard. We build, refine and validate such models and, importantly, state which parts are supported by the template and which are effectively invention.
Key Features & Advantages
State-Specific Modelling
Where a target adopts distinct active, inactive or ligand-bound states, templates are chosen for the state your project needs rather than for global sequence identity alone.
Serious Loop Treatment
Loops are rebuilt with conformational sampling and energy scoring, with alternative conformations delivered when the loop is genuinely mobile instead of a single arbitrary pose.
Multi-Template Assembly
Different regions can be modelled from different templates where that is better supported, with the boundaries and their justification documented.
Validated Before Delivery
Every model passes geometry, packing and environment checks, and the report identifies regions not reliable enough to support design decisions.
Where This Service Is Used
1Building a structure for a target where prediction alone is insufficient or unconvincing
2Producing a ligand-bound or specific conformational state for docking and design
3Modelling a close orthologue or mutant from an existing experimental structure
4Repairing experimental structures with missing loops, disordered termini or poorly resolved side chains
5Generating consistent structural models across a protein family for comparative analysis
Methods and Techniques Applied
1Template identification by sequence and profile search, with coverage, identity and conformational state assessed together
2Alignment construction and manual correction in functionally critical regions, where automated alignment often fails
3Single- or multi-template model building with restraint-based or fragment-assembly methods
4Loop sampling and refinement with energy scoring, including alternative conformations for mobile loops
5Side-chain rebuilding, hydrogen placement, protonation assignment and restrained energy minimisation
6Validation with Ramachandran, rotamer, clash and residue-environment analysis, plus comparison with the template
What We Need From You
1Target sequence in FASTA, with the isoform, construct boundaries and any tags specified
2The conformational state required and any ligand or cofactor that should be present
3Preferred templates if your team already has a view on which structure is relevant
4The downstream application, which determines how aggressively loops and side chains are refined
What You Receive
1Final model in PDB format together with the alignment and template record
2Alternative loop conformations where the loop cannot be resolved to a single pose
3Validation report covering geometry, packing and per-region reliability
4Prepared structure ready for docking or MD, with protonation states documented
5Written assessment stating which regions may carry design decisions and which may not
What We Compute and Report
1Template coverage, sequence identity and alignment quality per modelled region
2Full-atom coordinates with alternative conformations for loops that cannot be resolved to one pose
3Per-residue model quality estimates distinguishing design-ready regions from unreliable ones
4Geometry and packing validation statistics for the finished model
5Binding site quality assessment reported separately from global model quality
Questions We Are Asked Before Scoping
Should we use homology modelling or AI prediction?
▸Usually prediction first, then template-based refinement toward the state you need. When a close template exists in the required conformation, template-based modelling is often the more faithful route. We choose during the feasibility review and explain why.
How low can template identity go?
▸Models below roughly 30% identity are useful for fold-level reasoning but not for side-chain-level design. We will build one and label it accordingly rather than present it as design-ready.
Can the model be used for virtual screening?
▸Often yes, provided the binding site region is well supported by the template. We assess site quality separately from global model quality, because that is what governs screening outcomes.
Interpretation and Limitations
Computational predictions prioritise hypotheses; they do not independently establish binding, biological activity, therapeutic efficacy, safety or clinical validity. Performance depends on input quality, target class, the structural evidence available and the applicability domain of the selected methods. Scores used for ranking are comparative devices and should not be read as measured quantities. All deliverables are for research use only and are not intended for diagnostic or therapeutic use.
Is This the Right Service for You?
Choose this when
A close template exists in the specific conformational or ligand-bound state you need
An experimental structure has missing loops, termini or poorly resolved side chains
Consistent models are needed across a protein family for comparative work
Consider another service when
No suitable template exists and the fold itself must be predicted
AI Protein Structure Prediction The loop is genuinely mobile and its dynamics matter more than one pose
Classical MD Simulation What Goes Wrong, and How We Handle It
These are the failure modes that produce confident-looking results which do not survive the next experiment. Each one is designed against rather than discovered afterwards.
Template chosen on global identity while modelling the wrong functional state
Templates are selected for the state the project needs, and the state of each candidate template is stated in the report.
A single arbitrary loop conformation presented as the answer
Mobile loops are delivered as ensembles with their energies, so design decisions can be tested for robustness across them.
Alignment errors in functionally critical regions propagating into the model
Alignments are corrected manually around active sites and binding regions rather than accepted from automated output.
A Worked Example
Rebuilding an active-site loop missing from the deposited structure
The situation
The only structure of the target had eleven disordered residues directly over the catalytic site, making docking impossible.
What the work involved
Templates with the loop resolved were identified, the alignment was corrected by hand in that region, and the loop was rebuilt with conformational sampling and energy scoring.
What came back
Three distinct loop conformations, two of which left the site accessible. Docking was run against both, and the poses that were consistent across them were prioritised.
Illustrative of the shape this work takes. Details are generalised, no client or target is identified, and scope and outcome vary with the system in question.
What Published Work Reports
The results below were published by independent groups and are cited to show what these computational approaches have achieved in the literature. They are not ICDMO project results, and outcomes vary substantially between targets, datasets and programmes. We do not promise comparable results for any particular project.
Around three quarters of predicted ligand poses fall within 2 Å
Current deep-learning structure prediction, evaluated on an independent protein–ligand benchmark, placed roughly 76% of ligands within 2 Å of the experimental pose, and reported approximately 80% by coordinate error — a substantial improvement over classical docking baselines on the same set, achieved without supplying a binding site. The corollary matters as much as the headline: roughly a quarter of cases still fall outside that threshold, which is why we report confidence per interface rather than delivering a single model.
Abramson et al., Nature, 2024 ↗Service Workflow
1
Scientific Consultation
We discuss the objective, the decision it supports and the data you hold. Nothing is quoted before we understand what the result has to answer.
2
Feasibility Review
We check that the data, target class and available structural evidence can support the proposed approach, and say plainly where they cannot.
3
Technical Proposal
Scope, methods, deliverables, timeline and price are confirmed in writing, together with the limitations we expect to report.
4
Computation and QC
The analysis runs on pinned tool versions in a containerised environment, with quality checkpoints and method comparison where that is informative.
5
Report and Follow-Up
You receive the ranked results, the documented methods, the limitations and a proposed validation plan, plus support in interpreting them.
Service Process
1
Online Consultation
›
2
Solution Matching
›
3
Service Contract
›
4
Project Execution
›
5
Data Delivery
Specifications & Turnaround
Note: ICDMO delivers computational (in silico) services. Results prioritise hypotheses for testing; they do not by themselves establish binding, biological activity, efficacy or safety. All deliverables are for research use only and are not intended for diagnostic or therapeutic use. Downstream experimental work is scoped separately and executed through qualified partners.
Get a Custom Quote
Our scientific team responds within 24 hours with a detailed technical proposal and pricing tailored to your project.
Contact Us Online ConsultationStandard Deliverables
✓Sequence-verified final product
✓Full QC documentation
✓Certificate of Analysis (CoA)
✓Electronic data files
✓Project completion report
Why ICDMO?
Computational CDMO Focus
Reproducible, Versioned Pipelines
Confidence & Limitations Reported
Experiment-Aware Recommendations
Response Within One Business Day