Skip to content

gosset schema ​

The published field list for each dataset: what a record can contain, and which names --fields and --sort accept.

bash
gosset schema              # every dataset, with its field count
gosset schema abstracts    # the fields, with a description each
$ gosset schema
Published schemas (schema_version v3):

  abstracts    31 fields
  companies    27 fields
  deals        29 fields
  drugs        98 fields
  events       21 fields
  news         32 fields
  results      42 fields
  targets      11 fields
  trials       73 fields

Run `gosset schema <entity>` for the field list.
$ gosset schema abstracts
abstracts: 31 fields

  abstract_id         Stable Gosset identifier for this presentation
  title               Presentation title as published by the meeting
  url                 Link to the abstract on the meeting or journal site
  doi                 DOI of the published abstract, where the meeting assigns one
  conference          Meeting series, e.g. ASCO, ESMO, ASH. Filter with `conference`
  conference_year     Year of the edition, e.g. 2025
  edition             Meeting edition as one label, e.g. "ASCO 2025"
  ...

Counts grow as fields are added; the command always shows the current list. The CLI has commands for drugs, trials, companies, deals, news and abstracts. events, results and targets are published for the API and SDK.

Why you'd use it ​

To find the right name for --fields and --sort. The API checks every field name. A name it doesn't publish is an error, never a silently empty column:

bash
gosset schema drugs | grep -i phase
  phase                       Furthest phase reached
  stage_evidence              Why `phase` reads what it does: stage_basis (registry_declared | …
  preclinical_substage        The preclinical stage a cited source states, on drugs whose phase is …
bash
gosset drugs --target PD-1 --fields name,phase,developers --table

Because older field names changed. In the current API, latest_phase is phase, display_name is name, symbol is ticker, and route_of_administration is route. An old name fails with a pointer to the new one:

$ gosset drugs --target PD-1 --fields name,latest_phase
{"error": "API error (400): unknown field(s) ['latest_phase'] for drugs. 'latest_phase' is published as 'phase'. See GET /v3/schema/drugs."}

The schema comes from the API rather than a copy inside the CLI, so it can't fall out of date with the data.

Notes ​

  • Default JSON output returns every published field that has a value. Empty fields are left out of each record. --fields narrows the output; it can't add fields.
  • --table shows a few columns per dataset so the table fits a terminal. That choice is not the schema; gosset schema is.
  • --include-ids adds id fields (such as target_ids and disease_class_ids) for joining records across datasets.
  • Filter names (for --where) are a separate list; see Query API.

Gosset Documentation