How do you find a pharma company's clinical pipeline?

Search ClinicalTrials.gov by sponsor. Registration is a legal and ethical requirement before a trial enrols patients, so the registry is the closest thing to a complete public pipeline: what is being tested, in which phase, on how many participants, for which condition, and who is collaborating. Searching by sponsor rather than by drug also surfaces trials where your target is a partner rather than the lead — often the more interesting result.

What is the difference between a trial sponsor and a collaborator?

The lead sponsor holds regulatory responsibility for the trial; collaborators are partner organisations contributing funding, a compound or capability. The distinction matters for competitive intelligence because partnerships often appear in the registry before either company announces them. A search for one company will surface trials where it appears only as a collaborator, so you need a field that tells you which side of the relationship your search term is on.

Where does US research funding data come from?

Two federal agencies publish separately. NIH RePORTER covers biomedical research and gives project title, funding institute, award amount, fiscal year and named principal investigators. NSF Award Search covers general science and engineering, and unusually also publishes the awardee organisation's phone number — measured at 98% fill in our runs, making it the only open award source with a direct contact route.

Can you see which drugs a company has had approved?

Yes. FDA publishes approved drug products with their sponsor, and separately registers medical device manufacturers and their listed devices. Approvals are the end of the pipeline; trials are the middle; grants are the beginning. Reading all three against the same company gives a complete lifecycle view that no single registry provides.

Why do many trials have no collaborators listed?

Because many trials are genuinely run by a single sponsor. In our measured runs the collaborators field populates on about 47% of rows — the field is empty precisely when there is no partner, which is information rather than a gap. The same applies to phase: observational studies carry no phase, which is why that field sits at roughly 98% rather than 100%.

Home › Guides › Pharma pipeline data

Mapping a pharma company from public records: grants, trials and approvals

A drug's life leaves a public trail in three separate registries — research funding at the start, clinical trials in the middle, FDA approval at the end. Each is open data, each is searchable by company, and almost nobody reads them together. This guide covers what each registry actually contains, where the fields go quiet and why, and how to assemble a company-level view of a pipeline.

The three stages, and the registry for each

Stage	Registry	What it tells you
Discovery & preclinical	NIH RePORTER, NSF Award Search	Who is funding which research, at which institution, with which investigator
Clinical development	ClinicalTrials.gov	What is being tested, in which phase, on how many participants, with which partners
Market	FDA Drugs@FDA, FDA device registrations	What was approved, for whom, and who manufactures it

Read alone, each answers a narrow question. Read together against one company, they show the shape of a business: where its science came from, what it is betting on now, and what it has already brought to market.

Clinical trials: search by sponsor, not by drug

Registration on ClinicalTrials.gov is required before a trial enrols patients, which makes it the closest thing to a complete public pipeline. The instinct is to search by compound; searching by sponsor is more useful, because it returns the whole portfolio and — importantly — trials where your target appears as a collaborator rather than the lead.

That distinction is where partnership intelligence lives. A real row from our measured run:

{
  "companyName": "Vertex Pharmaceuticals Incorporated",
  "awardId": "NCT05668741",
  "awardTitle": "A Phase 1/2 Study of VX-522 in Participants With Cystic Fibrosis (CF)",
  "counterparty": "Moderna, Inc",
  "status": "ACTIVE_NOT_RECRUITING",
  "startDate": "2023-02-27",
  "endDate": "2026-04-21"
}

This row surfaced under a search for Moderna — which is not the lead sponsor. Vertex is. Moderna appears as the collaborator supplying the mRNA platform. A tool that only returns "trials sponsored by X" would have missed the relationship entirely, which is why the dataset carries a flag stating which side of the trial your search term sits on.

Where the data goes quiet, and why: in our runs phase populates on ~98% of rows (observational studies have no phase), collaborators on ~47% (many trials genuinely have a single sponsor), and enrollment on 100%. Sparse fields here are usually the registry being precise rather than incomplete.

Research funding: two agencies, two very different records

NIH RePORTER is the biomedical record. Each row is a project award: title, funding institute, award amount, fiscal year, and the named principal investigators. Continuing projects are funded yearly, so one project appears as several rows — a core project number groups them, and totalling a programme without it will double-count badly.

{
  "companyName": "REGENERON PHARMACEUTICALS, INC.",
  "awardId": "1U01HG004085-01",
  "awardTitle": "Production of Targeted Null Mutations for 10,000 Genes…",
  "counterparty": "National Human Genome Research Institute",
  "amount": 5311300,
  "fiscalYear": 2006
}

NSF Award Search covers general science and engineering. It matters here for an unusual reason: NSF publishes the awardee organisation's phone number, which came back on 98% of rows in our measurement. Among open award registries that is rare — most give you an organisation name and nothing to act on.

One warning for anyone parsing NSF directly: it ships dates as US MM/DD/YYYY. Parsing them as dd/mm/yyyy silently swaps day and month and produces impossible values like month 13 or 26. We hit exactly this — 39 of 51 rows carried invalid dates before the fix.

Approvals: the end of the pipeline

FDA publishes approved drug products with their sponsors, and separately maintains device manufacturer registrations with listed devices. For competitive work these answer the question the other registries cannot: what actually made it. Pair an approval list with a trial list for the same company and the attrition between them is the story.

Assembling the company view

Four practical notes from doing this at scale:

Match on legal name, not brand. Registries use the registered entity — REGENERON PHARMACEUTICALS, INC., not "Regeneron". Searching the brand alone silently under-returns.
Resolve institutions to a stable ID. Universities and hospitals appear under dozens of name variants across these registries. The ROR registry gives every research organisation a canonical ID that all three datasets can join on.
Expect the same project several times. Yearly NIH increments and trial amendments both create repeated rows. Group before you sum.
Don't treat sparse as missing. No collaborator means a solo trial. No phase means an observational study. Encoding those as nulls-to-be-fixed produces wrong analysis.

Which dataset for which question

Question	Dataset
A company's full clinical pipeline, by phase	Clinical Trials by Sponsor
Which drugs a company has had approved	FDA Drug Approvals by Sponsor
Medical device manufacturers and registrations	FDA Device Manufacturers
NIH grants at an organisation, with investigators	NIH Grant Data by Organization
NSF awards, with awardee phone numbers	NSF Award Data by Company
Resolving institution names to one canonical ID	Research Organization Registry
Nonprofit research foundations and their finances	US Nonprofit Data (Form 990)

Where this connects to the rest of healthcare data

Pipeline data answers what a company is developing. Two adjacent questions need different registries:

What has it already brought to market? FDA publishes approved products with the sponsor company — see the US healthcare organisation guide, which also covers device manufacturer registrations.
Who is the legal entity, and who owns it? Pharma group structures are complex and cross-border, and a sponsor name on a trial is frequently a subsidiary. The LEI and SEC filings guide covers resolving that properly.

The pairing worth building is trials against approvals, per sponsor. Trials show intent; approvals show conversion. Neither number alone says much, and the relationship between them over time is a genuine measure of a company's ability to execute rather than merely to announce.

Is this data free to use?

All of it is published by public bodies precisely so that it can be read. Trial registration exists so patients and clinicians can see what is being tested. NIH and NSF publish awards so taxpayers can see where research money goes. FDA publishes approvals as a matter of public safety. The records describe organisations, studies and funded projects — investigator names appear as the agencies themselves publish them, and no patient-level data is involved at any point.

Frequently asked questions

How do you find a company's clinical pipeline?

Search ClinicalTrials.gov by sponsor. Registration precedes enrolment, so the registry is the most complete public view of what a company is developing — including trials where it is a partner rather than the lead.

What's the difference between sponsor and collaborator?

The lead sponsor carries regulatory responsibility; collaborators contribute funding, a compound or a capability. Partnerships frequently appear in the registry before either party announces them.

Where does research funding data come from?

NIH RePORTER for biomedical research, NSF Award Search for general science. NSF additionally publishes the awardee's phone number — 98% fill in our measurements.

Why do many trials list no collaborators?

Because they have none. The field populates on about 47% of rows, and its absence means a single-sponsor trial rather than missing data.

Explore trials by sponsor → See all 87 datasets