Table of Contents
Lead optimisation services help pharmaceutical and biotechnology companies refine promising compounds before deciding whether any should progress towards preclinical candidate selection. The work can involve medicinal chemistry, biological testing, computational modelling, ADME and DMPK studies, pharmacokinetic assessment and early investigation of potential safety liabilities.
Choosing a drug discovery contract research organisation, or CRO, is therefore more complex than comparing prices or laboratory capacity. A suitable provider must understand the scientific problem, produce reliable data, communicate limitations and help the sponsor make defensible decisions.
There is no single CRO that is “top” for every programme. The right partner depends on the target, therapeutic area, chemical modality, development strategy, internal capabilities and risks associated with the starting compounds.
This guide explains what lead optimisation services cover, how the process works and how UK drug-discovery teams can assess a potential CRO.
Editorial note: This article provides general scientific and business information. It does not constitute regulatory, clinical, legal or investment advice. References to CRO service pages describe publicly stated capabilities and should not be interpreted as independent rankings or endorsements.
What Are Lead Optimisation Services?
Lead optimisation services are coordinated scientific activities used to improve the properties of promising lead compounds and determine whether one or more compounds are suitable for preclinical candidate consideration.
The process usually aims to balance several characteristics rather than maximise one measurement. These may include:
- Potency against the intended target
- Selectivity over unintended targets
- Solubility and permeability
- Chemical and metabolic stability
- Absorption, distribution, metabolism and excretion
- Pharmacokinetic exposure
- Early safety-related liabilities
- Synthetic accessibility
- Formulation feasibility
- Intellectual-property potential
Public service descriptions from Charles River Laboratories, Sygnature Discovery, Domainex and o2h Discovery illustrate how CROs may combine chemistry, biology, computational design and DMPK support during this stage. Their terminology and service boundaries differ, showing why buyers should evaluate the proposed programme rather than rely on a service label alone.
Lead optimisation does not prove that a compound is safe or effective in people. It remains an early stage of drug discovery, before formal non-clinical development and human clinical trials.
Where Does Lead Optimisation Fit in Drug Discovery?
A simplified small-molecule discovery pathway may include:
- Target identification and validation
- Hit identification
- Hit confirmation
- Hit-to-lead development
- Lead optimisation
- Preclinical candidate selection
- Non-clinical development
- Clinical development
The boundaries are not always fixed. A programme may return to an earlier stage when new data reveal an assay problem, an off-target effect, poor exposure or a difficult chemical liability.
Hit-to-lead Versus Lead Optimisation
Hit-to-lead work examines early chemical starting points and attempts to identify one or more credible lead series. Researchers may confirm activity, establish an initial structure–activity relationship and remove weak or misleading hits.
Lead optimisation starts with a more established compound or series. It applies repeated design and testing cycles to balance activity, selectivity, pharmacokinetics, safety-related properties and developability.
Domainex describes hit-to-lead as a process for focusing subsequent work on the most promising chemical series, while Sygnature highlights the use of integrated chemistry, biology, modelling and ADMET/DMPK work. These are first-party descriptions, but they demonstrate the multidisciplinary nature of the transition from hits to leads.
What is Candidate Selection?
Candidate selection is the formal decision to nominate a compound for further development against predefined criteria. The evidence package may include pharmacology, pharmacokinetics, toxicology-related findings, manufacturability and other programme-specific considerations.
A compound should not be selected merely because it has the highest potency. A less potent molecule may be preferable when it offers better selectivity, exposure, solubility or safety margins.
What Is the Main Goal of Lead Optimisation?
The main goal is to find an acceptable balance between biological performance and developability.
Drug discovery teams often face competing objectives. A structural change that improves potency may reduce solubility. A modification that slows metabolism may increase activity against an unwanted target. A compound that performs well in a biochemical assay may fail to enter cells or reach the relevant tissue.
Lead optimisation is therefore a multiparameter decision problem.
Potency and Target Engagement
Potency measures how much compound is needed to produce a defined effect in a particular test. However, assay context matters. Biochemical potency does not automatically establish cellular activity, target engagement in a living system or therapeutic relevance.
A CRO should explain:
- Which potency measurements are being used
- Whether the assay reflects the intended mechanism
- How controls and replicates are designed
- Whether cellular target engagement is required
- How results affect compound progression
Selectivity
A selective compound acts more strongly on the intended target than on other biological targets.
Insufficient selectivity can produce misleading efficacy signals or introduce unwanted biological effects. Selectivity panels should be chosen according to the target class, known pharmacology, intended indication and chemical series—not simply because a standard panel is available.
Physicochemical Properties
Physicochemical characteristics influence how a compound behaves in assays, formulations and biological systems.
Common measurements include:
- Molecular weight
- Lipophilicity
- LogP or LogD
- Aqueous solubility
- Permeability
- Ionisation
- Chemical stability
- Protein binding
No single value determines whether a compound is suitable. Results must be interpreted together and in the context of the intended route of administration and target tissue.
ADME and DMPK
ADME refers to absorption, distribution, metabolism and excretion. DMPK means drug metabolism and pharmacokinetics.
These disciplines help researchers understand how a compound enters the body, moves between tissues, changes through metabolism and is eventually removed. They may also identify potential interactions with enzymes or transporters.
The Charles River overview of ADME and DMPK studies lists examples such as solubility, permeability, plasma protein binding, metabolic stability, CYP interactions, transporter studies, tissue distribution and in vivo pharmacokinetics. The precise package should be selected for the compound and programme rather than applied as a fixed checklist.
Early Safety-related Assessment
Discovery-stage teams may investigate potential liabilities involving:
- Off-target pharmacology
- Cardiac ion channels
- Cytochrome P450 enzymes
- Reactive metabolites
- Cytotoxicity
- Genotoxicity indicators
- Organ-specific concerns
- Exposure relative to effective concentration
These screens can support early risk decisions, but they do not replace the formal non-clinical safety programme required before or during clinical development.
Which Lead Optimisation Services Can a CRO Provide?
The scope may range from a single specialist assay to a fully integrated programme.
Medicinal Chemistry
Medicinal chemists design and synthesise new compounds based on available biological, structural and pharmacokinetic data.
Typical responsibilities include:
- Structure–activity relationship analysis
- Scaffold and substituent modification
- Potency and selectivity optimisation
- Reduction of metabolic liabilities
- Improvement of solubility or permeability
- Assessment of synthetic feasibility
- Design of patentable compound series
- Preparation and characterisation of compounds
The o2h Discovery medicinal chemistry overview describes lead optimisation as the refinement of hits or leads to improve efficacy, selectivity, safety-related properties and patentability. As with any CRO claim, buyers should ask for programme-relevant evidence rather than assume that a general capability statement proves suitability.
Assay Development and Biological Testing
Reliable biology is essential because chemical optimisation depends on the quality of the test system.
Services may include:
- Biochemical assays
- Cell-based assays
- Functional assays
- Target-engagement studies
- Mechanism-of-action investigations
- Selectivity testing
- Biomarker assessment
- Disease-relevant models
A CRO should document assay controls, sources of variability, acceptance criteria and procedures for investigating unexpected results.
Structural and Computational Chemistry
Computational methods can help scientists prioritise compounds and explore possible relationships between molecular structure and biological behaviour.
Applications may include:
- Structure-based drug design
- Ligand-based drug design
- Docking
- Molecular dynamics
- Property prediction
- Virtual screening
- Compound prioritisation
- Retrosynthetic planning
Predictions remain hypotheses until they are tested experimentally.
In Vitro and in Vivo DMPK
In vitro testing can rapidly compare compounds across properties such as permeability, metabolic stability and enzyme interactions.
In vivo pharmacokinetic studies may then help researchers understand the combined effect of absorption, distribution, metabolism and excretion within a living system. The study must answer a defined scientific question and be designed with appropriate ethical consideration.
For UK organisations, the principles of Replacement, Reduction and Refinement should be considered wherever animals are used. The UK National Centre for the Replacement, Refinement and Reduction of Animals in Research explains that the 3Rs provide a framework for replacing animal use where possible, reducing the number used and refining procedures to minimise harm.
Formulation and Developability Support
A compound may need early formulation work to enable pharmacokinetic or pharmacology studies.
Relevant considerations include:
- Intended route of administration
- Solubility
- Dose concentration
- Stability
- Excipient compatibility
- Solid-state properties
- Scale-up feasibility
- Future manufacturing complexity
Addressing these factors early can prevent the team from advancing a compound that later proves impractical to formulate or manufacture.
How Does a Lead Optimisation Programme Work?
Most programmes use an iterative design–make–test–analyse, or DMTA, cycle.
1. Define the Target Candidate Profile
The sponsor and CRO should first agree what a successful compound needs to achieve.
Criteria may cover:
- Required potency
- Selectivity expectations
- Target exposure
- Route of administration
- Dosing frequency
- Pharmacokinetic behaviour
- Safety margins
- Formulation requirements
- Intellectual-property objectives
These criteria should guide decisions without becoming inflexible. New evidence may justify revising them.
2. Review the Starting Evidence
Before launching new experiments, the team should assess:
- Target validation
- Assay quality
- Existing compounds and data
- Known liabilities
- Available structural information
- Competitor molecules
- Patent position
- Translational assumptions
A detailed review may expose problems that would otherwise consume months of chemistry and testing.
3. Prioritise the Most Important Risks
Not every property should be optimised at once.
For example, a programme with excellent potency but rapid metabolic clearance may prioritise metabolic stability and exposure. A series with strong activity but poor selectivity may require early off-target investigation before additional resources are committed.
4. Run DMTA Cycles
Scientists design compounds, synthesise them, test them and analyse the resulting data.
Each cycle should answer specific questions:
- Did the structural change improve the intended property?
- Was another property harmed?
- Does the result support the working hypothesis?
- Should the series continue?
- What experiment would provide the greatest decision value next?
This approach aligns with the wider principle of data-driven decision-making: data should inform a meaningful operational choice rather than simply accumulate in a database.
5. Apply Progression and Stopping Criteria
A well-governed programme defines when to:
- Advance a compound
- Repeat a result
- Investigate an anomaly
- Pause a series
- Change strategy
- Terminate a series
- Nominate a candidate
Stopping a weak series can be a successful outcome when it prevents further investment in an unsuitable direction.
How Should a Company Choose a Top Drug Discovery CRO?
A top drug discovery CRO is the provider best suited to the specific programme not necessarily the largest, cheapest or most widely recognised organisation.
Match Expertise to the Scientific Problem
The sponsor should evaluate whether the proposed team has relevant experience with:
- The target class
- Therapeutic area
- Chemical modality
- Assay type
- Route of administration
- Tissue or biological barrier
- Known compound liabilities
- Required development stage
General experience in drug discovery is not the same as experience with the programme’s particular risks.
Assess the Scientific Strategy
A credible proposal should explain more than which services will be delivered.
It should identify:
- The central scientific questions
- The greatest uncertainties
- The sequence of experiments
- The rationale for each study
- Decision points
- Progression criteria
- Stopping criteria
- Alternative strategies
Buyers should be cautious when a proposal provides a long activity list without explaining how the results will guide decisions.
Review Interdisciplinary Integration
Integrated delivery can reduce hand-off delays and improve communication between chemistry, biology and DMPK teams. However, calling a service “integrated” does not prove that integration occurs in practice.
The sponsor should ask:
- Do teams share the same data environment?
- Who is accountable for cross-disciplinary decisions?
- How quickly are results reviewed?
- Are scientists able to challenge one another’s interpretations?
- How are disagreements documented?
- Will the sponsor have direct access to technical specialists?
Domainex and Sygnature both publicly emphasise multidisciplinary delivery, but buyers should verify the operating model through named personnel, governance arrangements and programme examples.
Investigate Data Quality and Reproducibility
The CRO should be able to explain:
- Assay validation
- Positive and negative controls
- Replicate strategy
- Acceptance criteria
- Instrument qualification
- Raw-data availability
- Audit trails
- Protocol deviations
- Data-review procedures
- Repeat-testing rules
A visually polished presentation is not a substitute for transparent experimental evidence.
Evaluate the Proposed Team
Sponsors should know who will perform and oversee the work.
Important questions include:
- Who is the scientific lead?
- Which specialists will be assigned?
- How much of their time is available?
- Will the sales-stage experts remain involved?
- How is staff turnover managed?
- Which services will be subcontracted?
- Who makes day-to-day decisions?
Examine Governance and Communication
Useful governance arrangements may include:
- Regular project meetings
- Written progress reports
- Decision logs
- Risk registers
- Milestone reviews
- Escalation routes
- Change-control procedures
- Budget tracking
- Defined sponsor and CRO responsibilities
Communication should make uncertainties visible rather than hide them.
Clarify Intellectual-property and Data Rights
The contract should address:
- Ownership of newly designed compounds
- Ownership of experimental data
- Background intellectual property
- Foreground intellectual property
- Invention reporting
- Patent-support responsibilities
- Rights to methods or platforms
- Publication restrictions
- Confidentiality
- Use of client data in computational models
These points are particularly important when AI or proprietary modelling platforms are involved.
Compare Commercial Models Carefully
Possible arrangements include:
- Fee-for-service
- Full-time-equivalent contracts
- Milestone-based programmes
- Integrated project agreements
- Hybrid structures
The lowest headline price may not represent the lowest total cost. Buyers should examine assumptions, project-management charges, pass-through expenses, shipping, subcontracting, repeat work and the process for changing scope.
Drug Discovery CRO Evaluation Scorecard
| Evaluation area | What to examine | Strong evidence | Warning sign |
|---|---|---|---|
| Scientific fit | Target, modality and therapeutic expertise | Comparable programme examples and relevant specialists | Generic claims covering every area |
| Strategy | Hypotheses, risks and decision criteria | Programme-specific experimental plan | List of activities without rationale |
| Biology | Assay relevance and reproducibility | Validation data, controls and acceptance criteria | Reliance on one unverified assay |
| Data quality | Raw data and review procedures | Traceable records and transparent deviations | Summary slides only |
| Integration | Chemistry, biology and DMPK coordination | Shared governance and rapid cross-team review | Unclear hand-offs |
| Team | Named scientists and availability | Accessible, stable technical leadership | Senior experts disappear after contracting |
| Communication | Reporting and escalation | Agreed meeting, reporting and decision structure | No accountable project lead |
| Intellectual property | Compound, invention and data ownership | Clear contractual provisions | Ambiguous foreground-IP terms |
| Commercial model | Scope, assumptions and change control | Transparent pricing and responsibilities | Unclear pass-through costs |
| AI capability | Data, validation and ownership | Relevant model evidence plus experimental testing | Unsupported claims of guaranteed acceleration |
The table should be used as a discussion framework, not a universal ranking formula.
Standalone Specialists or an Integrated CRO?
Neither model is automatically better.
A standalone specialist may be appropriate when the sponsor has strong internal project leadership and needs a narrow capability, such as structural biology, medicinal chemistry or a particular DMPK assay.
An integrated CRO may be valuable when several disciplines must work concurrently and the sponsor wants one governance structure.
Possible advantages of standalone specialists
- Deep expertise in a narrow field
- Independent interpretation
- Flexibility to choose a different provider for each discipline
- Reduced dependency on one organisation
Possible disadvantages
- More hand-offs
- Greater coordination burden
- Incompatible data formats
- Slower communication
- Fragmented accountability
Possible advantages of an integrated CRO
- Shared project management
- Faster circulation of results
- Easier coordination of DMTA cycles
- Fewer external hand-offs
- One contractual relationship
Possible disadvantages
- Vendor concentration
- Higher switching costs
- Limited independent challenge
- Dependence on the provider’s internal platforms
- Risk that “integration” exists mainly in marketing
The correct model depends on the sponsor’s internal expertise, resources and desired level of control.
How Is AI Used in Lead Optimisation?
Artificial intelligence and machine-learning tools may support compound prioritisation, property prediction, molecular design, pattern detection and synthetic planning.
They can help teams examine a large number of possibilities before deciding which compounds to make and test. However, AI-generated outputs do not independently prove biological activity, selectivity, pharmacokinetic performance, safety or clinical effectiveness.
The most credible AI-enabled workflow combines computational predictions with experimental validation and expert scientific judgement.
The same supplier-evaluation discipline used when assessing software development companies is relevant here: buyers need to understand the technology, the people operating it, the data behind it and the evidence that it performs reliably in the intended context.
Questions to Ask an AI-enabled CRO
A sponsor should ask:
- What data were used to train or calibrate the model?
- Is the model relevant to the programme’s chemical space?
- How is predictive uncertainty reported?
- How are recommendations experimentally validated?
- Can scientists explain why a compound was prioritised?
- How is confidential client data protected?
- Is client data used to improve shared models?
- Who owns AI-generated structures and associated data?
- How does the model perform against an appropriate baseline?
- What happens when the prediction and experimental result conflict?
A Sygnature article on keeping AI accountable from in silico work to in vivo testing similarly frames experimental evidence as an important part of evaluating computational predictions. It remains a provider-authored source, so its statements should be assessed alongside independent scientific evidence.
Practical Example: A UK Biotechnology Company Chooses a CRO
Consider a fictional UK biotechnology company with two promising small-molecule series.
Both series show cellular activity, but the company has limited internal medicinal chemistry capacity. One series has poor metabolic stability, while the other produces selectivity concerns.
A structured selection process could proceed as follows:
Define the Objective
The company establishes a target candidate profile covering potency, selectivity, exposure, dosing route and safety-related margins.
Identify the Central Risks
Instead of requesting a broad package of standard services, the company identifies metabolic stability and selectivity as the immediate decision risks.
Shortlist Relevant Providers
CROs are shortlisted according to target-class experience, medicinal chemistry strength, DMPK capability and evidence of resolving similar problems.
Request Programme-specific Proposals
Each CRO is asked to explain:
- Which risk should be addressed first
- What experiments are proposed
- How each result would change the strategy
- Which specialists will be assigned
- How long each decision cycle is expected to take
- What assumptions affect cost
Apply the Scorecard
The sponsor compares scientific fit, data quality, integration, team access, intellectual-property provisions and commercial flexibility.
Begin With a Controlled Work Package
Rather than immediately committing the full programme, the sponsor may start with a defined phase containing clear deliverables and decision criteria.
Review Results at Agreed Decision Gates
The programme continues only when evidence justifies the next stage.
This approach does not guarantee a successful candidate. It does, however, improve transparency and help the company make disciplined investment decisions.
UK Regulatory and Development Considerations
Discovery work should be planned with later development requirements in mind, but early CRO studies should not be presented as regulatory approval or proof of clinical performance.
The UK Medicines and Healthcare products Regulatory Agency states that organisations may request scientific advice at any stage of a medicine’s development or regulation. The usefulness of such advice depends on the questions and supporting documents provided, and it cannot account for future changes in scientific knowledge or requirements.
The appropriate timing for regulatory engagement depends on the programme. Sponsors should seek qualified advice where decisions may affect future non-clinical, clinical, quality or regulatory plans.
Common Mistakes When Outsourcing Lead Optimisation
Selecting Primarily on Price
A cheaper programme can become expensive when assays must be repeated, data are incomplete or the team pursues the wrong scientific question.
Focusing Only on Potency
High potency cannot compensate for every selectivity, exposure, stability or safety-related problem.
Starting Without Decision Criteria
Without progression and stopping rules, activity can continue without producing a defensible decision.
Accepting “Integrated” at Face Value
The sponsor should verify how disciplines share information and who is accountable for cross-functional decisions.
Overlooking Raw-data Access
Sponsors need enough information to review, reproduce and transfer the work.
Ignoring Intellectual-property Terms
Unclear ownership of compounds, inventions, data or AI-generated designs can create serious commercial risk.
Measuring Progress by Compound Count
The number of compounds synthesised is less important than the quality of the hypotheses tested and decisions enabled.
Treating AI Predictions as Experimental Evidence
Predictions should guide testing, not replace it.
Confusing Candidate Nomination With Development Success
A nominated candidate still faces substantial non-clinical, regulatory and clinical evaluation.
Questions to Ask Before Appointing a CRO
Before signing an agreement, a sponsor should ask:
- Which comparable programmes has the proposed team completed?
- Who will lead the project scientifically?
- Which programme risk should be addressed first?
- How will chemistry, biology and DMPK teams work together?
- What are the progression and stopping criteria?
- How will raw data and protocols be delivered?
- How are failed or inconclusive experiments reported?
- Which work will be subcontracted?
- Who owns compounds, data and inventions?
- How will confidential data be used in AI or predictive systems?
- How are scope and staffing changes managed?
- What evidence supports proposed timelines and costs?
- How will the CRO prepare an integrated candidate-selection package?
- What happens when the sponsor and CRO disagree on interpretation?
Key Takeaways
- Lead optimisation is a multiparameter process, not a potency contest.
- A top drug discovery CRO is the provider that best fits the programme’s scientific and operational needs.
- Relevant expertise, assay quality and decision strategy matter more than a long services list.
- Raw-data access, reproducibility and transparent reporting are essential.
- Integrated delivery can be useful, but the operating model must be verified.
- Intellectual-property and data-use terms should be settled before work begins.
- AI can help prioritise ideas, but its predictions require experimental validation.
- UK sponsors should consider future development requirements and seek current regulatory advice where appropriate.
Conclusion
Effective lead optimisation services combine medicinal chemistry, biology, computational methods, ADME/DMPK and disciplined programme governance.
The right CRO should demonstrate relevant expertise, transparent scientific reasoning, reliable data practices and contractual clarity. It should also be willing to identify uncertainty, recommend stopping weak work and explain how each experiment supports a decision.
Rather than asking which company is universally the top drug discovery CRO, sponsors should ask a more useful question: which provider is best equipped to resolve the specific scientific and commercial risks in this programme?
That question leads to a more defensible partnership and a better-informed candidate-selection process.
Frequently Asked Questions
What Are Lead Optimisation Services?
Lead optimisation services are scientific activities used to improve promising compounds and assess whether any should be considered for preclinical candidate selection. They may include medicinal chemistry, biological assays, computational modelling, ADME/DMPK, pharmacokinetics and early safety-related testing.
What is the Difference Between Hit-to-lead and Lead Optimisation?
Hit-to-lead work evaluates and improves early chemical hits to identify credible lead series. Lead optimisation begins with more established leads and attempts to balance potency, selectivity, pharmacokinetics, safety-related properties and developability.
How Long Does Lead Optimisation Take?
There is no universal timeline. Duration depends on the quality of the starting compounds, target complexity, assay readiness, required studies, chemistry cycle times and liabilities discovered during the programme. CRO estimates should be treated as programme-specific assumptions rather than guarantees.
What Does a Medicinal Chemistry Cro Do?
A medicinal chemistry CRO designs, synthesises and evaluates compounds to investigate structure–activity relationships and improve properties such as potency, selectivity, solubility, metabolic stability and synthetic feasibility.
Are Adme and Dmpk the Same?
They are closely related but not identical. ADME describes absorption, distribution, metabolism and excretion. DMPK focuses on drug metabolism and pharmacokinetics and helps explain the movement and transformation of a compound within biological systems.
What is a Preclinical Candidate?
A preclinical candidate is a compound formally selected for further non-clinical development against programme-specific criteria. Candidate selection does not establish that the compound will be safe or effective in humans.
Can AI Replace Medicinal Chemists?
AI can support compound design, property prediction and prioritisation, but it cannot independently validate biological activity, safety or development suitability. Medicinal chemists and other scientists must interpret predictions and test them experimentally.
Should a Company Use One Integrated Cro?
An integrated CRO may simplify coordination, but it is not always the best option. Sponsors with strong internal leadership may prefer specialist providers. The decision should consider scientific expertise, governance, data integration, independence and vendor concentration.
How Can a Sponsor Protect Its Intellectual Property?
The sponsor should agree clear contractual terms covering background IP, newly created compounds, inventions, experimental data, AI-generated designs, confidentiality, publications and patent support before the work begins.
How Should CRO Proposals Be Compared?
Proposals should be compared using scientific fit, strategy, team expertise, assay quality, raw-data access, governance, intellectual property, commercial assumptions and the evidence supporting capability claims—not price alone.
