What a passport is
A passport aggregates several tools into one profile. Here is the model behind it, how it differs from running tools yourself, and when to reach for one.
A passport is a multi-source aggregated report. Instead of running ADMET, synthesis, vendor, docking, literature, and safety tools one at a time and stitching the results together, you name a subject — a compound, a target, a drug, a query, a disease — and the passport orchestrates the data collection and presents everything in one interactive dashboard. This page explains the model behind passports so you know when one saves you work and what its numbers mean.
The one idea to hold onto: a passport is subject-keyed, not tool-keyed. You ask a question about one thing, and the passport answers it from every source that has data on that thing — then tells you, with a completeness score, how much it found.
The trade it makes
A canvas card shows the result of one tool. A passport shows the result of many, keyed to a single subject. That is the whole distinction: a card answers "what did this tool find?"; a passport answers "what do I need to know about this thing?"
You reach for a passport when the answer spans several tools and you want it assembled — a compound's properties and its synthesis routes and its vendors and its docking behavior, in one view with a completeness score. You reach for a single tool when you want to control one step or inspect an intermediate result. See Decisions for the choice.
The five types
Each passport type takes one kind of input and pulls from a fixed set of sources. The subject you name determines both.
| Passport | Input | Pulls from |
|---|---|---|
| Compound | 1–10 SMILES | ADMET, synthesis, vendors, docking, RDKit properties |
| Target | UniProt ID | UniProt, PDB, druggability, BioBERT literature |
| Safety | Drug name | FAERS, FDA labels, recalls, organ-risk assessment |
| Literature | Search query | BioBERT entities, relationships, trends, company activity |
| Therapeutic Area | Disease name | Target landscape, compounds, literature, safety |
The same generator also builds narrower reports on the same model: pathway, biomarker, drug repurposing, patent landscape, pharmacovigilance, and gene comparison.
How aggregation works
A passport is a fan-out, then a merge. You name the subject; the passport queries every source in its set — some are the same tools you could call in chat (ADMET, docking, BioBERT literature), some are external databases (FAERS for safety, PubChem and vendor catalogs for compounds) — and folds every result into one profile organized by section. The compound passport, for example, merges molecular properties, an ADMET panel, synthesis routes, vendor availability, and any docking results the platform already has for that structure into a single card.
A passport generates in one pass and returns as soon as its sections are assembled — it does not run as a long background job you poll. When you generate one inside a chat, it appears on your canvas automatically as a report card with expandable sections and export options; generated outside a chat, its data comes back inline.
The completeness score
Every passport reports a completeness score from 0 to 100%, weighted across its sections. A low score is information, not failure — it tells you which sources had no data for this subject, so you know where the profile is thin before you rely on it. A compound with no deposited docking poses and no vendor listings scores lower on those sections not because the passport failed, but because those sources genuinely returned nothing for that molecule.
Read the score as coverage, not quality: it measures how many of the passport's sources contributed, not how good any one result is. Check the score before you treat a passport as complete.
Who can generate which
Passport access depends on your plan.
- Compound and Target passports are available to every signed-in user.
- Literature passports require Grow or higher.
- Safety and Therapeutic Area passports require Flourish or higher.
Every passport can be exported as JSON, CSV, and Markdown; PDF and DOCX exports need Grow or higher. If you generate a passport you do not have access to, the platform tells you which plan it needs rather than returning a partial result.
Passport or tools?
Generate a passport when you want the assembled profile of one subject and are happy with the standard source set. Run tools individually when you want to pick a specific tool, tune its parameters, or inspect an intermediate before continuing. The two are complementary: a passport is often how you start on a subject, and a single tool is how you drill in once you know which section to chase.
Read next
Passports & Reports
The five types in detail — inputs, sources, durations, and access.
Canvas card vs. Passport
One tool's result versus a multi-source profile.
Which tool?
Match a single scientific goal to the tool that answers it.
How the orchestrator picks tools
How a plain-language goal becomes a chain of tools.
Was this page helpful?
Reading an in silico result against the bench
What a predicted structure, pose or property score has been shown to agree with experimentally, how that agreement is measured, and where the figures stop.
Concept clusters
Every platform concept, grouped into five clusters. Each cluster answers one question about how work moves through the platform.