Ref D4
Computational Service
Confidence Reported
Scoped After Feasibility Review
Service Overview
Computational safety profiling exists to remove the compounds most likely to fail late, and to do it while substitution is still cheap. We assess cardiac liability through hERG and related channels, genotoxicity through Ames and structural alerts, hepatotoxicity risk, sensitisation potential and assay-interfering behaviour such as aggregation and PAINS substructures. Every alert is delivered with the matching substructure and the mechanistic reasoning, because an unexplained red flag is not actionable.
Key Features & Advantages
Alerts With Mechanism Attached
Each flag identifies the substructure and the mechanistic basis, which allows the team to judge whether the alert is relevant to their chemistry or a known false positive.
Cardiac Safety Prioritised
hERG liability receives detailed treatment, including the structural features driving it, because cardiac toxicity remains one of the most common causes of late-stage attrition.
Assay Interference Screened
PAINS, aggregators, redox cyclers and fluorescence interferers are identified so that screening effort is not spent chasing artefacts.
Mitigation Suggested
Where a liability is structurally localised, we propose modifications known to reduce it, rather than leaving the team with a flag and no route forward.
Where This Service Is Used
1Triaging screening hits to remove compounds with obvious safety liabilities
2Assessing a lead series for cardiac, genotoxic or hepatic risk before candidate selection
3Explaining an unexpected toxicity signal observed in an in vitro or in vivo study
4Supporting an internal go/no-go decision with a documented computational safety assessment
5Screening a library before purchase to avoid acquiring problematic chemotypes
Methods and Techniques Applied
1hERG and cardiac ion channel liability prediction with structural feature attribution
2Ames mutagenicity and genotoxicity prediction combined with expert structural alert systems
3Drug-induced liver injury risk assessment and hepatotoxicity alert screening
4Skin and respiratory sensitisation prediction where the route of exposure makes it relevant
5PAINS, aggregation, redox cycling and assay interference substructure screening
6Consolidated risk profile per compound with alerts ranked by severity and confidence
What We Need From You
1Compounds as SMILES or SDF with stereochemistry and salt form specified
2Intended indication, patient population and route of administration, which affect what risk is acceptable
3Any measured toxicology or safety pharmacology data for calibration and context
4Known liabilities in the chemical series, so the analysis can focus where it is most useful
What You Receive
1Consolidated safety profile table with predicted endpoints and confidence per compound
2Structural alerts with the matching substructure highlighted and the mechanism explained
3Cardiac liability assessment with the features responsible identified
4Assay interference flags separating genuine liabilities from screening artefacts
5Risk-ranked compound list with mitigation suggestions where the liability is structurally addressable
What We Compute and Report
1Cardiac ion channel liability estimate with the structural features driving it
2Mutagenicity and genotoxicity predictions combined with expert structural alerts
3Hepatotoxicity risk indication and sensitisation potential where the route makes it relevant
4Assay interference flags: promiscuous substructures, aggregation and redox-cycling risk
5Consolidated per-compound risk profile with alerts ranked by severity and confidence
Questions We Are Asked Before Scoping
Can computational profiling replace toxicology studies?
▸No. It prioritises and de-risks; regulatory safety assessment requires the specified in vitro and in vivo studies. These deliverables are for research use only.
How many alerts are false positives?
▸A meaningful proportion. Alert systems are deliberately conservative. Because we report the mechanism and matching substructure, your team can judge each case rather than discard compounds automatically.
Is a clean prediction a clean compound?
▸No. Absence of an alert means no recognised liability pattern was detected, not that the compound is safe. Novel chemotypes are precisely where computational safety models are weakest.
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
Screening hits need triaging before obvious liabilities consume assay capacity
A lead series is approaching candidate selection and needs a documented safety assessment
An unexpected toxicity signal has appeared and a structural explanation is needed
Consider another service when
The compound fails on exposure rather than on safety
Physicochemical Property Prediction Reactive metabolite formation is the specific concern
Metabolic Stability & CYP Prediction 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.
Conservative alert systems discarding viable chemistry
Every alert is delivered with the matching substructure and its mechanism, so your team can judge relevance instead of deleting compounds automatically.
A clean prediction read as a safe compound
Absence of an alert means no recognised liability pattern was found — which is weakest exactly where the chemistry is novel, and we say so.
Assay artefacts pursued as real activity
Aggregation, redox cycling and fluorescence interference are screened so they are excluded before, not after, the follow-up assay.
A Worked Example
Cardiac liability across a chemical series
The situation
One compound in a series showed hERG activity and the team needed to know whether the whole scaffold was affected.
What the work involved
Cardiac liability was predicted across every analogue with structural feature attribution, and the results were mapped onto the structural differences within the series.
What came back
The liability tracked a basic centre and a lipophilic tail rather than the core. Three modifications were proposed that retained potency vectors while lowering predicted risk, and the series was not abandoned.
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.
Property prediction is accurate enough to triage, not to replace measurement
Across the published literature, lipophilicity and surface-property predictions are generally reliable, while aqueous solubility models typically carry errors around one log unit and absolute human pharmacokinetic prediction remains weak. We therefore report predictions with their expected error and applicability-domain status, and recommend measurement at the point where a decision turns on the number rather than the rank order.
Method comparison literature, ongoing ↗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