/ writing / pharma

Drug intel is scattered across three federal systems. One actor shouldn't be.

2026-08-27 · cynix · 4 min read

FDA labeling lives in openFDA. Trials live in ClinicalTrials.gov. Gene context lives in MedlinePlus / NCBI. Three systems, three response shapes, three "read the docs" rabbit holes — and none of them talk to each other. You just want: "what's the trial + label + gene picture for pembrolizumab?"

Pharma / Biotech Intelligence Pack fuses all three into one actor with one flat record shape. Search by drug, add a gene, get a unified typed dataset.

Sources & what you get

SourceWhat you get
FDA (openFDA)Drug labeling: brand/generic/ingredient match, indications, warnings, adverse events
ClinicalTrials.gov v2Studies by condition/intervention: title, recruitment status, phase, conditions
MedlinePlus Genetics (NCBI Gene)Gene summary linked to MedlinePlus for a symbol like BRCA1

Example output

{
  "query": "pembrolizumab + BRCA1",
  "source": "ClinicalTrials.gov",
  "title": "A Study to Investigate the Safety of Pembrolizumab",
  "status": "RECRUITING",
  "phase": "PHASE1",
  "detail": "Neoplasms",
  "sourceUrl": "https://clinicaltrials.gov",
  "attribution": "ClinicalTrials.gov (NLM, public domain)"
}

Every run also emits a RUN_SUMMARY row reporting per-source hit counts, so an empty result is explicit ("no matches") rather than a silent source failure.

Why this over the websites?

Use cases

Try it: Pharma / Biotech Pack on Apify — three sources, one schema, keyless, schedulable.