Ref G1
Computational Service
Confidence Reported
Scoped After Feasibility Review
Service Overview
An integrated SBDD programme is not a single calculation but a running design–make–test–analyse loop in which computation sets the agenda for each round. We hold the computational side: predicting and modelling the target, designing and prioritising compounds, calculating binding energies, and then re-analysing once your assay and structural data return. Cycles typically run two to four weeks, and each one begins with what the previous round actually showed rather than with what was predicted.
Key Features & Advantages
Structured Cycles, Not Ad Hoc Requests
Each round has a defined question, a defined compound set and defined success criteria, so progress can be judged rather than assumed.
Experimental Data Fed Back Every Round
Assay results and new structures are incorporated as they arrive, and models that failed to predict them are corrected — which is where most of the learning happens.
Method Matched To Decision Value
Docking for triage, MM/GBSA for ranking, FEP where a synthesis decision is expensive. Compute is spent where it changes an outcome.
One Team Across The Loop
The same scientists handle structure, design and analysis, so context is not lost between handoffs and prediction errors are traced rather than forgotten.
Where This Service Is Used
1Running hit-to-lead optimisation where structural information is available or obtainable
2Lead optimisation programmes needing sustained computational support across many cycles
3Supporting an internal chemistry team that has no in-house computational chemistry function
4Reviving a stalled series by re-examining the structural basis of its SAR
5Building the structural rationale required for a candidate selection dossier
Methods and Techniques Applied
1Target structure establishment by experimental data, prediction or modelling, with conformational states characterised
2Binding site and hot-spot analysis, and construction of the structure-based design hypothesis
3Compound design by manual structure-based reasoning, generative methods or analogue enumeration
4Prioritisation cascade: docking, MM/GBSA and FEP applied according to decision cost
5Property and liability prediction integrated into selection rather than applied at the end
6Retrospective analysis of each round, with model correction where predictions failed
What We Need From You
1Target structure or sequence, plus any co-crystal structures your team holds
2Current series with measured activity, selectivity and property data
3Programme objectives and the criteria a candidate must eventually meet
4Synthesis and assay capacity per cycle, which sets the size of each design round
What You Receive
1Prioritised compound designs each cycle with structural rationale and predicted profiles
2Updated binding mode models and SAR interpretation as new data arrives
3Free energy calculations for the decisions that warrant them
4Cycle reports covering what was predicted, what was observed and what changed as a result
5Cumulative programme documentation suitable for internal review and candidate dossiers
What We Compute and Report
1Ranked designs per cycle with predicted binding, property and liability profiles
2Free energy estimates for the transformations that justify the calculation cost
3Retrospective accuracy per cycle: what was predicted against what was measured
4Updated binding mode models as new structural and assay data arrives
5Cumulative structure–activity interpretation across the programme
Questions We Are Asked Before Scoping
Do you perform the synthesis and assays?
▸No. ICDMO is a computational CDMO. We design, prioritise and analyse; synthesis and testing are executed by your team or a qualified partner, and we help plan and interpret that work.
How many cycles before we see progress?
▸It depends on starting point and target tractability. Most programmes see the modelling become genuinely predictive after two or three cycles of feedback, which is when the loop starts paying for itself.
What if predictions keep failing?
▸Then we say so and diagnose why — wrong binding mode, unmodelled flexibility, an unrecognised assay artefact. A model that cannot explain observed SAR is reported as such rather than quietly retuned.
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 chemistry team is running cycles and has no in-house computational chemistry function
Structural data exists but is not being converted into design decisions
A programme needs sustained support over months rather than a single calculation
Consider another service when
A single structure needs interpreting, not a programme
Protein–Ligand Co-Crystal Structure Analysis The programme has not yet found chemical matter
High-Throughput Virtual Screening 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.
Expensive methods applied to decisions that do not need them
Method is matched to decision cost: docking triages, end-point methods rank, alchemical calculations decide.
Models quietly retuned until they agree with the latest data
Each cycle reports what was predicted before the data arrived, and a model that cannot explain observed SAR is reported as failing.
Design rounds that test many variations of one idea
Each round is constructed so the compounds discriminate between distinct hypotheses.
A Worked Example
Six cycles of hit-to-lead support
The situation
A biotech had a confirmed hit, a co-crystal structure and a chemistry team, but no computational support.
What the work involved
Each cycle: analyse new structures and assay data, design and prioritise compounds, apply free energy calculations to the decisions that warranted them, and report prediction against observation.
What came back
By the third cycle the model reproduced the measured rank order of new compounds well enough to be used for selection. The programme reached its potency criterion in six cycles with 84 compounds synthesised.
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.
Structure prediction now exceeds classical docking on independent benchmarks
On a held-out protein–ligand benchmark, deep-learning complex prediction placed roughly three quarters of ligands within 2 Å of the crystallographic pose, outperforming established docking programs on the same set. For design programmes this changes what can be attempted before a co-crystal structure exists — while leaving binding-mode certainty, and therefore the most consequential chemistry decisions, dependent on experimental structures.
Abramson et al., Nature, 2024 ↗Free energy methods can resolve the differences that decide a synthesis
Published benchmarks on congeneric series report errors near 1 kcal/mol, sufficient to rank analogues that differ by around tenfold in potency. In an integrated design loop this is the calculation reserved for decisions where making the wrong compound costs more than the computation.
Protocol and benchmark studies, Scientific Reports, 2019 ↗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