Ref K2
Computational Service
Confidence Reported
Scoped After Feasibility Review
Service Overview
Stability engineering is one of the more reliable things computation does for proteins. We identify destabilising positions, predict the effect of substitutions on folding free energy, combine consensus, structural and evolutionary evidence, and deliver a ranked mutation set together with modelled combination variants. Function is protected explicitly: active site, interface and dynamically important residues are excluded or treated with care, because a stable protein that no longer works is not an improvement.
Key Features & Advantages
Several Independent Prediction Routes
Physics-based free energy calculations, consensus and evolutionary analysis, and machine-learning predictors are combined; mutations supported by several methods have markedly better success rates.
Function Protected By Design
Catalytic, binding and allosterically coupled residues are identified and excluded, so stabilisation does not silently cost activity.
Combinations Modelled For Additivity
Stabilising mutations are combined and modelled rather than simply stacked, since epistasis frequently makes the combined effect smaller than the sum.
Failure Mode Targeted Specifically
Thermal unfolding, aggregation, pH sensitivity and oxidation are different problems with different fixes, and designs target the one that actually limits your protein.
Where This Service Is Used
1Raising the melting temperature of an industrial or therapeutic enzyme
2Improving shelf life and formulation stability of a protein product
3Reducing aggregation during expression, purification or storage
4Stabilising a protein for structural studies or as a vaccine antigen
5Increasing tolerance to organic solvents, extreme pH or oxidative conditions
Methods and Techniques Applied
1Structure preparation and analysis of packing, cavities, surface charge and flexible regions
2Computational saturation mutagenesis with folding free energy prediction at every position
3Consensus analysis across a curated homologue alignment, including thermophilic relatives where they exist
4Machine-learning and language-model scoring of substitution plausibility
5Functional site protection: exclusion of catalytic, binding and dynamically coupled positions
6Combination design with explicit modelling and molecular dynamics of the leading multi-mutants
What We Need From You
1Protein sequence and structure, experimental or modelled
2The stability problem being solved, with any measured melting temperature or aggregation data
3Functional constraints: activity that must be retained, binding partners, required conditions
4How many variants you can express and characterise
What You Receive
1Ranked single mutations with predicted stability effects from each method used
2Designed combination variants with modelled structures and predicted cumulative effects
3Functional risk assessment per mutation, flagging those near sensitive regions
4Structural rationale explaining what each mutation is intended to do
5Recommended testing order, designed so early results inform later rounds
What We Compute and Report
1Predicted folding free energy change for every substitution at every position
2Consensus and evolutionary scores from family analysis, reported alongside the energetics
3Agreement between independent prediction routes, used to rank confidence
4Predicted cumulative effect of combination variants, modelled rather than summed
5Functional risk score per mutation based on proximity to catalytic and binding residues
Questions We Are Asked Before Scoping
How much stabilisation is achievable?
▸Successful single mutations commonly contribute a few degrees each, and combinations can add up substantially. Outcomes vary by protein, and a fraction of predicted mutations will not work as predicted.
Will stabilisation reduce activity?
▸It can, since flexibility often matters for function. That is why functional sites are protected and designs are delivered with their functional risk flagged, so activity can be measured alongside stability.
Do we need an experimental structure?
▸A good model is usually sufficient for surface and packing analysis, though experimental structures improve accuracy. Model confidence is carried through to the per-mutation assessment.
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
An enzyme or protein product unfolds at process or storage temperature
Aggregation during expression, purification or formulation limits yield
A protein must be stabilised for structural study or as a vaccine antigen
Consider another service when
Catalytic activity rather than stability is the limiting property
Enzyme Design & Directed Evolution The protein sits in a membrane
Membrane Protein 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.
Stabilising mutations that silently remove activity
Catalytic, binding and dynamically coupled residues are identified and excluded, and every design carries a functional risk flag.
Single-method predictions acted on directly
Physics-based, consensus and learned approaches are combined, and mutations supported by all three are prioritised.
Combination variants assumed to be additive
Multi-mutants are modelled and simulated, because stacked stabilising mutations frequently underdeliver.
A Worked Example
Raising the melting temperature of a process enzyme
The situation
The enzyme lost activity above 45 °C, limiting reaction temperature and throughput.
What the work involved
Full-sequence saturation scanning with folding free energy prediction, consensus analysis against thermophilic homologues, exclusion of active-site and coupled positions, and modelling of combination variants.
What came back
Thirty-two single mutations ranked by cross-method agreement and eight modelled combinations, each with its predicted functional risk, ordered so early results informed the later rounds.
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.
Machine-learning-guided evolution raised a reaction yield from 12% to 93%
An active-learning workflow optimising five interacting active-site residues improved the yield of a desired non-native cyclopropanation product from 12% to 93% in three rounds of laboratory experimentation. The design principle is the one we apply: computation chooses which small set of variants to make, and measured results retrain the model between rounds.
Nature Communications, 2025 ↗1.6- to 42-fold activity improvements from model-selected enzyme variants
A machine-learning-guided enzyme engineering campaign evaluated 1,217 variants across 10,953 reactions to build predictive models for amide synthetases. Model-selected variants showed 1.6- to 42-fold improved activity relative to the parent enzyme across nine pharmaceutical targets, illustrating both the scale of data that makes models predictive and the spread of outcomes that remains.
Nature Communications, 2024 ↗De novo binder design succeeds at rates that vary enormously by target
Recent design methods report nanomolar binder hit rates between roughly 17% and 82% across evaluated targets, while independent evaluations of earlier pipelines on other targets report designs failing on expression, non-specific binding or undetectable affinity. The honest reading is that de novo binder design now works, that success is strongly target-dependent, and that panels must be sized accordingly rather than on a best-case figure.
Comparative design studies, bioRxiv, 2025 ↗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