Ref I4
Computational Service
Confidence Reported
Scoped After Feasibility Review
Service Overview
In silico affinity maturation narrows the mutational space before any library is built. We model the antibody–antigen complex, perform computational saturation mutagenesis across the CDRs and contact positions, score each substitution for its predicted effect on binding and stability, and combine favourable mutations into a ranked variant panel. The practical benefit is focus: tens of designed variants to express and measure instead of a library of millions to screen.
Key Features & Advantages
Binding And Stability Scored Together
Mutations are assessed for both affinity and folding stability, because a large fraction of affinity-improving substitutions destabilise the domain.
Orthogonal Scoring Methods
Physics-based energy calculations and language-model or statistical approaches are combined, and mutations favoured by both receive higher confidence than those favoured by one.
Combinations Modelled, Not Assumed
Beneficial single mutations are not simply summed; combinations are modelled to detect the epistasis that frequently makes stacked mutations underperform.
Liabilities Excluded By Design
Substitutions that introduce deamidation, isomerisation, oxidation or glycosylation motifs are filtered out, so the improved variant does not arrive with a new developability problem.
Where This Service Is Used
1Improving the affinity of a lead antibody before candidate selection
2Designing a focused mutagenesis library informed by computation rather than random diversity
3Rescuing affinity lost during humanisation or format conversion
4Fine-tuning specificity to reduce cross-reactivity with a close off-target
5Identifying the CDR positions that actually carry the binding energy
Methods and Techniques Applied
1Antibody–antigen complex modelling, or preparation of an experimental structure where one exists
2Interface analysis identifying contact residues and the energetic contribution of each
3Computational saturation mutagenesis across CDR and contact positions with energy-based scoring
4Machine-learning and language-model scoring for evolutionary plausibility and fitness
5Stability assessment of every candidate mutation alongside its predicted binding effect
6Combinatorial design with explicit modelling of the combined variants and liability filtering
What We Need From You
1Antibody sequences and the antigen sequence or structure
2A complex structure if one exists; otherwise complex modelling is included in the scope
3Measured affinity for the parent and for any variants already tested, which calibrates the ranking
4The improvement target and the number of variants you can express and measure
What You Receive
1Position-by-position mutational scan with predicted effects on binding and stability
2Ranked single and combination variants with the rationale for each
3Structural models showing what each mutation is intended to change at the interface
4Liability screening results confirming no new developability risk was introduced
5Recommended testing order, designed so that the results discriminate between hypotheses
What We Compute and Report
1Position-by-position mutational scan with predicted binding and stability effects
2Interface energy decomposition identifying which contacts carry the binding
3Combination variant predictions that model epistasis rather than summing single effects
4Evolutionary plausibility scores for every substitution considered
5Liability screening results confirming no new developability risk was introduced
Questions We Are Asked Before Scoping
How much affinity improvement can we expect?
▸Successful campaigns commonly achieve a few-fold to roughly an order of magnitude from computational design. Larger improvements happen but are not predictable, and we will not promise a number.
Is computational maturation better than display?
▸It is complementary. Display explores far more sequence space; computation focuses effort and explains the result. Many teams use computation to design a smarter library rather than to replace selection.
What if no complex structure exists?
▸We model it, and predictions are then less reliable. Where epitope information is also absent, we recommend epitope mapping first, since maturation without knowing the interface is largely guesswork.
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 lead antibody needs measurably better affinity before candidate selection
Affinity lost during humanisation or format conversion has to be recovered
A focused mutagenesis library should be designed rather than randomised
Consider another service when
Developability rather than affinity is the limiting problem
Full-Length IgG Antibody Design A library will be built and selected experimentally
Antibody Library Design 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.
Affinity-improving mutations that destabilise the domain
Every substitution is scored for folding stability alongside binding, and the trade-off is reported per mutation.
Beneficial single mutations assumed to add up
Combinations are modelled explicitly, because epistasis routinely makes stacked mutations underperform the sum of their parts.
A better binder arriving with a new developability liability
Substitutions that introduce deamidation, isomerisation, oxidation or glycosylation motifs are filtered out during design.
A Worked Example
Recovering affinity lost in humanisation
The situation
The humanised construct had lost roughly twenty-fold affinity relative to the murine parent.
What the work involved
The complex was modelled for both parent and humanised forms, the interface was decomposed per residue, and framework and CDR positions were scanned for substitutions that restored the lost contacts.
What came back
Two Vernier-zone positions accounted for most of the loss. A graded panel of sixteen variants was delivered, spanning full recovery with lower humanness through to partial recovery with germline frameworks retained.
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.
10- to 140-fold affinity improvements from computational design
Computational interface redesign produced a tenfold affinity improvement in cetuximab, reaching 52 pM, and a 140-fold improvement to 30 pM for an anti-lysozyme model antibody — improvements achieved by designing a small number of interface substitutions rather than by library selection.
Lippow, Wittrup & Tidor, Nature Biotechnology, 2007 ↗Interface models predict which substitutions help, with measurable accuracy
A 2024 study combining structural interface analysis with sequence-derived evolutionary information reported a predictive model for antibody–antigen interaction with an area under the curve of 0.83 and precision of 0.89 on held-out data, and experimentally confirmed a designed point mutation that improved affinity 2.5-fold to 2 nM. Typical computational maturation gains are modest per mutation, which is why a designed panel is measured rather than a single variant.
Experimental validation study, 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