Pillar 1 of 6 · Rationale
Should this drug actually work?
The scientific case for the asset, rebuilt from public target, pharmacology and structural evidence — so the biology is pressure-tested before a phase is spent on it.
Should this drug work?
What it answers
Should this drug work?
Before design, before sites, before budget: is there a defensible scientific reason to expect this molecule to move this disease? This pillar answers that from the public record — the same record a reviewer, a partner or an investment committee will check you against.
Is the target real?
Whether the target–disease link is carried by human genetics and clinical precedent, or rests mostly on expression and text-mining. Those are not the same claim, and the assessment does not let them read the same.
Does your molecule engage it?
What has actually been measured against the resolved target — and, separately, what has only been measured against other molecules in the class. Class evidence is never presented as your compound binding.
What breaks first?
The off-target, drug–interaction and safety liabilities that the target, the class and the molecule already carry in public data — and whether the target is structurally tractable at all.
Inside the pillar
4 analyses, each one named and cited
These are the analyses that actually run — not a category list. Each states what it computes and what it needs, because an analysis that quietly degrades is worse than one that reports it could not run.
Target validation
Scores the target–disease association across Open Targets' evidence channels — genetic association, somatic mutation, known drug, clinical, RNA expression, literature, animal model and affected pathway — and bands the result on the strength and breadth of the evidence, weighting human genetics above expression and text-mining.
Needs
The study target and its indication. The target is resolved through a synonym and alias resolver rather than a string match; when it cannot be matched confidently, no molecule-specific headline is claimed and the indication landscape is not passed off as target validation.
Binding affinity
Pulls measured ligand affinities for the resolved target from BindingDB and the IUPHAR/BPS Guide to Pharmacology, and separates rows attributable to your molecule from rows measured against the class. Bands are computed over human, non-censored rows only; non-human and bounded values (>, <, ~) are counted separately and never binned into a potency band.
Needs
A named molecule and a resolvable target accession. Sponsor-held affinity data can be overlaid on the public rows, with each row labelled public or internal.
Off-target & interaction risk
Combines the target's known safety liabilities and expression profile, the marketed class's US label precedent (boxed warnings, warnings and precautions, interaction text, CYP and transporter tags from FDA structured product labels), class-level FAERS reaction profiles, and the molecule's own measured secondary-pharmacology activities against a hERG / 5-HT2B / muscarinic / CYP / transporter checklist.
Needs
The molecule, its class and the resolved target. Lanes that did not run — a retired upstream source, an unlicensed commercial interaction database — report as not run, with the checklist coverage stated. Not measured is never rendered as clean.
Structural viability
Lists experimental co-crystal structures for the target from PDBe with their ligand interactions and resolution distribution, and reports Open Targets tractability buckets for the modality you are pursuing. Predicted structures are consulted only when there are no experimental entries, and are labelled as predicted.
Needs
A resolved target accession. Sponsor-internal crystallography rows are folded in on top of the public set when you have them.
Where the evidence comes from
Public sources, named
Everything in this pillar comes from open scientific and regulatory records. Two lanes are licence-gated and stay dark unless you hold the licence — a commercial drug-interaction database and the European pharmacovigilance feed — and the tile says which lanes ran rather than reporting their absence as a clean result.
Open Targets
Open Targets Consortium · Quarterly
Target–disease association evidence integrating genetics, pathways, and known drugs — feeds PTRS scoring.
opentargets.org (opens in a new tab)ChEMBL
EMBL-EBI · Quarterly
Manually curated bioactivity database of drug-like molecules — used to characterize molecule risk and mechanism precedent.
ebi.ac.uk (opens in a new tab)PubMed
NCBI / U.S. National Library of Medicine · Daily
Biomedical literature index of 36M+ citations — used to ground indication baselines and standard-of-care evidence.
pubmed.ncbi.nlm.nih.gov (opens in a new tab)FDA
U.S. Food & Drug Administration · Weekly
Approvals, labels, guidance, and adverse-event signals (openFDA) used to anchor regulatory-uplift and safety risk.
open.fda.gov (opens in a new tab)Also read by this pillar
What you get
What lands in the assessment
Every association score, affinity row, label excerpt and structure carries the public record it came from and the date it was read.
- A Rationale tab in the assessment with one tile per analysis, each carrying its own confidence band and the records it was built from
- Affinity rows split into your molecule, the class, and the censored or non-human rows that were excluded from the bands
- A safety and interaction panel that names the lanes that ran and the coverage of the ones that did
Read by these expert roles
A role drafts from this pillar; a human reviews and signs off. The sign-off is a 21 CFR Part 11 e-signature on a hash-chained record.
How the roles workHonest limits
What this pillar cannot tell you
Every assessment method has a boundary. Publishing ours is the point — a number you cannot check is worth less than a gap you can see.
It cannot tell you your molecule works.
Public evidence can say a target is well-validated and the chemistry is plausible. It is not a substitute for your own preclinical package, and it says nothing about formulation, human pharmacokinetics, or whether you can manufacture the thing.
A missing measurement is not a clean result.
If a secondary-pharmacology panel was never published, or an interaction source is not licensed, the tile reports the lane as not run and states how much of the checklist has any measurement at all. It never converts silence into safety.
A nameless asset reads at class level.
For a pre-clinical compound with no public identity, target and class evidence is what exists. The assessment gives you the class precedent and labels it as such — you do not get compound-specific affinity or safety data that was never measured.
Tractability is a bucket, not a plan.
Structures and tractability buckets say the target has been drugged, or has not. They do not tell you that your chemistry series can reach it, or hold it selectively enough to dose.
The rest of the assessment
No pillar decides alone
Each pillar answers one question. The Verdict composes all six into a single go/no-go, with the risks that could still change it.
See a verdict you can actually check.
Send us a protocol — or just a molecule and an indication. We'll return a fully cited feasibility assessment you can trace, line by line, back to public data — yours to defend in a bid, take to your board or investment committee, or hand to a regulator.