ProductUse casesPricingDocs
日本語Request access

Data sources

Effective: December 1, 2026

Besides our company database, AGcollectio uses public clinical trial registries, public data from regulators and governments, bibliographic details of conference abstracts and more. This page sets out, for each source, what we use it for, its terms of use, and how the source is shown in reports.

About third-party data

Most of the data listed here is created and published by organizations other than us. When we load it, we process it (extracting, classifying, matching names and so on), but we do not guarantee that the original data is accurate, complete or up to date. Errors or gaps in the original data appear in reports as they are.

When you use tables or charts from a report in materials outside your company, keep the source notes attached to them and follow the conditions for each dataset on this page. This is the same approach as the "Data sources" article of the Terms of Service.

Companies and pipeline assets

Our company database

  • Use: biotech companies, pipeline assets, technologies, partnerships and company tags, and disease primers (definition, patient numbers, standard of care and unmet needs).
  • Content: data we have collected and organized ourselves, based on company announcements and other sources. Disease primers give a source for each sentence.
  • In reports: as this is our own data, tables carry no source line.

Crunchbase (funding)

  • Use: companies' total funding.
  • How it is shown: funding based on Crunchbase data is shown as a band (a range), not as the amount itself.
  • In reports: the note on tables that show funding says that funding is based on Crunchbase data, with a link to Crunchbase.

Clinical trials

ClinicalTrials.gov (AACT)

  • Use: industry trials on ClinicalTrials.gov, the US clinical trial registry. We use them for trial populations, design, posted results, serious adverse events and eligibility criteria. We load the monthly releases of AACT, the copy published by the Clinical Trials Transformation Initiative (CTTI), and extract and classify the data ourselves.
  • Terms of use: within the US, the data is in the public domain (not covered by copyright). Outside the US, copyright may apply, and anyone distributing the data is asked to state that it comes from ClinicalTrials.gov, the date of the data and how it was changed (ClinicalTrials.gov terms and conditions).
  • In reports: tables and charts that use this data carry a line below them that names ClinicalTrials.gov (AACT) with the data date and says that we extracted and classified the data. The date is the date of the data we loaded.

EU Clinical Trials Register (EudraCT)

  • Use: the EU Clinical Trials Register of the European Medicines Agency (EMA), mainly trials filed in Europe up to 2022. We use it to count European trials, and match it against the US registry to show differences between development in the US and in Europe.
  • Terms of use: the EMA allows copying and distribution, including for commercial purposes, on condition that the EMA is acknowledged as the source and that this acknowledgment is included in each copy (EMA legal notice).
  • In reports: the notes of sections that use this data include text that names the EMA and the register, gives the register's URL, and says that we processed the data.

EU Clinical Trials Information System (CTIS)

  • Use: trials filed in Europe from 2023 onwards. CTIS succeeds EudraCT, but because filings overlap, we do not add the two together and use CTIS as a separate dataset.
  • Terms of use: public data from the EMA, which can be reused if the source is acknowledged (the same EMA legal notice as above).
  • In reports: the notes of sections that use this data include text that names the EMA and the system, gives the system's URL, and says that we processed the data.

Conference abstracts

  • Use: abstracts from the American Society of Clinical Oncology (ASCO) Annual Meeting and its Gastrointestinal Cancers and Genitourinary Cancers Symposia, the American Association for Cancer Research (AACR) Annual Meeting, the International Conference on Molecular Targets and Cancer Therapeutics held by the AACR, the US National Cancer Institute and the European Organisation for Research and Treatment of Cancer, the American Society of Hematology (ASH), the San Antonio Breast Cancer Symposium (SABCS), the Alzheimer's Association International Conference (AAIC) and the American Diabetes Association (ADA). We get the bibliographic details of each abstract (conference, year, abstract number, title, DOI and affiliations) from Crossref, and for some abstracts we extract the trial result figures given in the abstract. Affiliations are used to link abstracts to companies.
  • Handling: we do not store abstract text or author names, and they do not appear in reports. Only the bibliographic details and the extracted figures appear in reports.
  • In reports: the notes of sections that use this data include, for each conference, text that names the conference and the journal the abstracts are published in, and says that each abstract's source is its DOI, that we got the details from Crossref and processed them, and that we do not hold abstract text or author names.

Approvals and drug prices

EMA medicines list (European approvals)

  • Use: the list of medicines published by the EMA (medicines with a European public assessment report, or EPAR). We use the European approval dates and therapeutic areas for approvals by indication and for differences between US and European approvals. European product names on drug cards also come from this list. We do not store the indication text as it is; we take out only the disease names.
  • Terms of use: can be reused if the source is acknowledged (the same EMA legal notice as above).
  • In reports: the notes of sections that use European approvals show the EMA as the source.

FDA (US approvals)

  • Use: we load Drugs@FDA from the US Food and Drug Administration (FDA) through openFDA's bulk download files, and use it for US approval dates and indications. We also use the FDA's list of biological products (the Purple Book) and the labels and approval letters the FDA publishes as leads for approval dates and disease names. We do not store the indication text as it is; we take out only the disease names.
  • Terms of use: openFDA data is provided under CC0 (public data whose rights have been waived), with no requirement to credit the source (openFDA license).
  • In reports: the notes of sections that use approvals say that US approvals are based on the FDA's Drugs@FDA and the Purple Book. On drug cards, the US product name is shown as coming from FDA approval records.

MED-RT (US drug and disease relationships)

  • Use: through RxClass from the US National Library of Medicine (NLM), we get MED-RT (Medication Reference Terminology) may_treat relationships, which link a drug to the diseases it is used to treat. Diseases are given as MeSH identifiers, and we use them as leads for disease names in US approvals. They are not approval records themselves.

NHI drug price list (Ministry of Health, Labour and Welfare)

  • Use: Japanese drug prices (the official prices under National Health Insurance) and whether each product is an originator or a generic.
  • Terms of use: data on the website of the Ministry of Health, Labour and Welfare can be used, including for commercial purposes, if the source is shown, in line with the Public Data License (Version 1.0). If the data has been processed, that must also be shown (the Ministry's terms of use (in Japanese), Public Data License (Version 1.0) (in Japanese)).
  • In reports: the notes of sections that use drug prices include one sentence that gives the name and URL of the Ministry's page and says that we processed the list.

NDB Open Data (Ministry of Health, Labour and Welfare)

  • Use: from the 11th NDB Open Data (aggregated data on prescription drugs from the National Database of Health Insurance Claims and Specific Health Checkups of Japan), we use only the names of therapeutic categories.
  • Terms of use: the same as for the NHI drug price list.
  • In reports: the notes of sections that use therapeutic category names include text that gives the name and URL of the source page and says that we processed the data.

Patient numbers and prevalence

Patient Survey (Ministry of Health, Labour and Welfare, e-Stat)

  • Use: estimated total patient numbers by disease in Japan. We use tables published on e-Stat, the portal site for Japanese government statistics.
  • Terms of use: can be used, including for commercial purposes, under the e-Stat terms of use. We show the source and that we processed the data (e-Stat terms of use (in Japanese)).
  • In reports: the notes of sections that use patient numbers show the source and that we processed the data.

Orphanet (rare disease prevalence)

  • Use: the prevalence of rare diseases (such as rates per 100,000 people). We load the data Orphanet publishes (Orphadata) and link it to our disease names through the MeSH mappings that Orphanet maintains.
  • Terms of use: CC BY 4.0 (Creative Commons Attribution 4.0). The data can be used, including for commercial purposes, if the source is shown, a link to the license is given and any changes are indicated. When the data is provided to third parties through a service that uses AI, Orphanet asks that the service state that Orphanet data was used and that the results may not accurately reflect the original data (Orphadata legal notice).
  • About our output: the prevalence tables in reports are Orphanet data processed by us, and the report text is written by AI. Either may not accurately reflect the Orphanet data.
  • In reports: the notes of sections that use prevalence say that the data comes from Orphanet, give the license name and link, and say that we processed the data.

Dictionaries and identifiers used to match names

MeSH (Medical Subject Headings, US National Library of Medicine)

  • Use: disease names, and the classification that groups diseases by area. We load the disease terms attached to ClinicalTrials.gov records and the MeSH files published by the US National Library of Medicine (NLM). The therapeutic areas in the EMA medicines list are also MeSH terms.

MedDRA (Medical Dictionary for Regulatory Activities, ICH)

  • Use: a terminology used in the regulation of medicines, covering adverse events and more. We use MedDRA terms written in registrations and filings in the following two places.
    • Serious adverse event tables: on screen and in PDF reports, the adverse event terms and system organ classes registered on ClinicalTrials.gov are shown as registered. We do not move them within the hierarchy or group them into higher-level terms.
    • Disease names for European trials: MedDRA disease terms written in EudraCT filings are used as leads to link the trials to our MeSH disease names. What appears in reports is the MeSH disease name.
  • When data goes outside: when you save a table to Excel or hand a report to another AI, tables that contain MedDRA terms are not included. When you export a TPP (target product profile) draft table from a discussion, the serious adverse event evidence under the safety item is not included. Both can still be seen on screen, and reports also in PDF.

Open Targets

  • Use: we use target (gene) synonyms only to recognize the same target. They do not appear in reports.
  • Terms of use: provided under CC0 1.0, with no requirement to credit the source (Open Targets license).

ChEMBL, UNII, PubChem and other drug identifiers

  • Use: to recognize the same drug, our drug list (the drug master) holds identifiers such as ChEMBL, UNII and PubChem. They do not appear in reports. On drug cards, the UNII is shown as one of the names.
  • Terms of use: ChEMBL data is provided under CC BY-SA 3.0 (Creative Commons Attribution-ShareAlike 3.0) (About ChEMBL).

Web search sources

  • Use: when questions, discussions and drug cards look things up on the web, AI searches public pages and attaches the source URLs to its answers.
  • Handling: source pages belong to third parties, and we have not checked whether their content is correct. Open the source page and check it before you use it.

Contact

If you have questions about how sources are handled, or find errors in the data, please let us know at contact@pangaea.tokyo.

Operated by Pangaea, Inc., Tokyo.

日本語

Product

  • How it works
  • Comprehensive data
  • Analysis frameworks
  • Correct data
  • Use cases
  • Pricing

Docs

  • All docs
  • Getting started
  • Requests and plan approval
  • Reading a report
  • Questions and rebuilding sections
  • Scheduled updates
  • Data sources and coverage
  • Pricing and usage
  • How your input is handled
  • FAQ

Legal

  • Terms of Service
  • Privacy Policy
  • Notice under the Act on Specified Commercial Transactions
  • Data sources

Company

  • Operated byPangaea, Inc.
  • Company website