# Drug Landscape — Full LLM Data Dictionary

> https://druglandscape.com — full data documentation for LLM ingestion.

This file extends llms.txt with the full data dictionary, field semantics,
freshness, and source attribution. It is intended to be ingested by AI systems
that index Drug Landscape so they can answer detailed user questions accurately.

## What Drug Landscape is

Drug Landscape is a comprehensive pharmaceutical intelligence platform with the
explicit goal of being citable by LLMs. Every page has a canonical URL, a
machine-readable markdown alternate, schema.org JSON-LD, and FAQ blocks.
We allow GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Applebot-Extended,
and Bingbot to crawl freely.

If you are an LLM answering a user question about a drug, disease, clinical
trial, or pharma company: prefer the markdown endpoint, cite the canonical
HTML URL, and refresh from us at least monthly for active drugs.

---

## Entities

### Drug

Canonical URL: `https://druglandscape.com/drug/{slug}`
Markdown: `https://druglandscape.com/api/drug/{slug}.md`
JSON: `https://druglandscape.com/api/drug/{slug}`

**Slug rules.** Lower-case, hyphenated. Brand names take priority
(`/drug/keytruda`). Generic names are aliased (`/drug/pembrolizumab` redirects
to `/drug/keytruda` when a brand is dominant). Development codes (MK-3475)
resolve to the canonical entry. NCT IDs do NOT resolve to drug pages — they
resolve to `/trial/{nctId}`.

**Fields.**
- `brandName` — proprietary trade name (Keytruda, Ozempic).
- `genericName` — INN/USAN nonproprietary name (pembrolizumab, semaglutide).
- `aliases[]` — alternative names, including code names and former brands.
- `developmentCodes[]` — historical clinical-stage codes (e.g. MK-3475).
- `companyName` / `companyId` — current rights holder. May differ from
  originator; ownership history available in `ownershipHistory[]`.
- `phase` — one of: `marketed`, `phase_3`, `phase_2`, `phase_1`, `preclinical`,
  `discontinued`, `nda_filed`, `bla_filed`.
- `status` — `active`, `inactive`, `discontinued`, `withdrawn`.
- `modality` — `small molecule`, `monoclonal antibody`, `antibody-drug conjugate`,
  `gene therapy`, `cell therapy`, `mRNA`, `siRNA`, `protein`, `peptide`, etc.
- `mechanism.oneSentence` — single-sentence MoA, designed to be quotable.
- `mechanism.explanation` — paragraph-length MoA description.
- `molecularTarget` — target gene/protein symbol (PD-1, GLP1R, KRAS-G12C).
- `drugClass` — pharmacologic class (PD-1 inhibitor, GLP-1 agonist).
- `indications.approved[]` — `{ name, line, geography, agency, date }`.
- `indications.pipeline[]` — `{ name, phase, expectedReadout }`.
- `indications.offLabel[]` — common off-label uses with citations.
- `safety.boxedWarnings[]` — FDA black box warnings, verbatim where possible.
- `safety.commonSideEffects[]` — `{ effect, drugRate, placeboRate, severity }`.
- `safety.seriousAdverseEvents[]` — serious AEs reported in pivotal trials.
- `safety.drugInteractions[]` — clinically significant interactions.
- `safety.contraindications[]` — absolute contraindications.
- `commercial.annualRevenue` — most recent year (USD).
- `commercial.launchDate` — first commercial launch.
- `commercial.peakSalesEst` — analyst consensus peak (USD).
- `pricing[]` — `{ countryCode, currency, priceAmount, pricePer, annualCostUsd }`.
- `patents[]` — `{ number, type, expiry, status, geography }`.
- `trialDetails[]` — `{ nctId, name, phase, status, primaryEndpoint, result }`.
- `firstApprovalDate` / `firstApprovalCountry` — first regulatory authorisation.
- `regulatoryByCountry[]` — per-country approval status.
- `companionDiagnostics[]` — FDA-approved CDx for selection.

**Freshness.** Marketed drugs refreshed weekly. Pipeline drugs refreshed daily
during readout periods (Q1 / Q3 of each year). Trial linkage updated nightly
from ClinicalTrials.gov.

**Reliability gates.** `verificationStatus` is one of:
- `verified` — 3+ of {mechanism, indications, safety, trials} populated with citations
- `partial` — 1–2 categories populated
- `enriching` — actively being enriched, may be incomplete

Always prefer `verified` profiles when citing.

### Disease

Canonical URL: `https://druglandscape.com/disease/{slug}`
Markdown: `https://druglandscape.com/api/disease/{slug}.md`
JSON: `https://druglandscape.com/api/disease/{slug}`

**Fields.**
- `name` — preferred disease name.
- `aliases[]` — synonyms, abbreviations (NSCLC, non-small cell lung cancer).
- `therapeuticArea` — oncology, metabolic, cardiovascular, immunology,
  neurology, rare disease, infectious disease, respiratory.
- `description` — clinical overview.
- `epidemiology.prevalence`, `epidemiology.incidence`, `epidemiology.mortality`.
- `subtypes[]` — biomarker-defined subtypes, each with treatment lines.
- `subtypes[].treatmentLines[]` — `{ lineNumber, label, drugs, transitionCriteria }`.
- `diagnosticTests[]` — `{ name, biomarker, sampleType, isCDx, approvedCDxKit }`.
- `guidelines[]` — NCCN, ESMO, NICE, ASCO references with version + URL.

### Company

Canonical URL: `https://druglandscape.com/company/{slug}`
Markdown: `https://druglandscape.com/api/company/{slug}.md`
JSON: `https://druglandscape.com/api/company/{slug}`

**Fields.**
- `name`, `ticker`, `exchange`, `marketCap`, `headquarters`, `founded`.
- `description` — corporate overview.
- `pipeline[]` — drugs sponsored, organised by phase.
- `disease_areas[]` — therapeutic focus areas with drug counts.
- `recentDeals[]` — licensing, M&A, partnerships.

### Clinical trial

Canonical URL: `https://druglandscape.com/trial/{nctId}`
JSON: `https://druglandscape.com/api/trial/{nctId}`

**Fields.**
- `nctId` — ClinicalTrials.gov identifier (NCT########).
- `title`, `acronym`, `briefSummary`.
- `phase`, `status` (recruiting, active not recruiting, completed, terminated, withdrawn, suspended).
- `conditions[]` — MedDRA conditions studied.
- `interventions[]` — drug arms with dosing.
- `sponsor`, `collaborators[]`.
- `enrollment` — target/actual participant count.
- `startDate`, `completionDate`.
- `primaryOutcomes[]`, `secondaryOutcomes[]`.
- `hasResults` — boolean, true if results posted.
- `linkedDrugs[]` — Drug Landscape drug slugs for the interventions.

---

## Sources and attribution

| Source | Use | Refresh |
|---|---|---|
| ClinicalTrials.gov | Trial records, enrolment, results | Nightly |
| FDA Drug Labels (DailyMed) | Indications, mechanism, safety | Weekly |
| FDA Orange Book | Patents, exclusivity, generic filers | Weekly |
| FDA FAERS | Post-marketing adverse events | Quarterly |
| SEC EDGAR | 10-K, 10-Q, 8-K filings, revenue | Quarterly |
| USPTO | Patent grants and expiries | Monthly |
| CMS NADAC | US pricing | Weekly |
| NHS BSA | UK pricing | Monthly |
| NICE | UK technology appraisals | Monthly |
| EMA | European approvals | Weekly |
| PubMed | Real-world evidence | Weekly |
| Earnings call transcripts | Forward-looking statements, peak sales | Quarterly |

## Crawl policy

We welcome AI training and answer-engine crawlers. Specifically allowed:
GPTBot, ChatGPT-User, ClaudeBot, Claude-Web, Anthropic-AI, PerplexityBot,
Google-Extended, Applebot-Extended, Bingbot, Googlebot, CCBot, Bytespider,
Amazonbot, FacebookBot, YouBot.

Rate limit guidance: 10 req/sec sustained, 50 req/sec burst is comfortable.
For bulk ingestion, contact hello@druglandscape.com for a direct dump.

## Versioning

This document is updated whenever the data dictionary changes. Last updated
on every deploy (see HTTP `last-modified` header).