Sources & roadmap
Every number on screen is fetched live from a public API and links back to its source record. There is no mock, sample or made-up data in the app or its code. If a source fails, that view shows the error and nothing else.
Data sources
| View | Source | What we take | Status |
|---|---|---|---|
| Trial news, landscape, investigators | ClinicalTrials.gov API v2 | Interventional Ph1–3 trials updated in the last 16 months; phases, status, dates, sponsors, interventions, overall officials | Live |
| Papers, investigator output | PubMed E-utilities | Trial-type papers in the last 60 days; author + indication counts | Live |
| Targets with genetic evidence | Open Targets Platform GraphQL (EMBL-EBI) | Disease→target associations with datatype scores (germline genetic association, somatic, drugs, pathways, literature…), tractability, clinical candidates, gnomAD constraint | Live |
| ASCO 2026 abstracts | Crossref (JCO 44:16_suppl; LBAs in 17_suppl) | Abstract number, title, authors, DOI | Live |
| AACR 2026 abstracts | Crossref (Cancer Research 86:7_Supplement) | Abstract number, title, authors, DOI | Live |
| SITC 2026 abstracts | Crossref (JITC 14 Suppl 2 and 3, open access) | Abstract number, title, authors, DOI | Live |
| ESMO 2026 abstracts | Parallel web research over published congress reports | Number, title, author, session, trial, NCT ID, drugs, source link; each row verified against a source page | Verified pull |
| ASH 2026 abstracts | Parallel web research (regular abstracts; LBAs embargoed to 7 Dec) | Same fields, same verification | Verified pull |
| Paediatric target flag | FDA Pediatric Relevant Molecular Target List | Gene symbols as the FDA lists them | Reference file |
| Actionable biomarkers | CIViC GraphQL (CC0) | Accepted evidence items: molecular profile, level, therapy, significance, citation | Live |
| Grants | NIH RePORTER API v2 | Projects by PI name or by grant number on a trial record | Live |
How events are derived
- Phase advance: an asset's first trial at a higher phase (Ph2 or later) than any earlier trial of the same asset in the indication, posted in the last 120 days.
- Results posted: the registry's results-first-posted date falls in the window.
- Readout due: a Ph2+ trial's primary completion date falls in the window. This flags an expected readout, not a disclosure.
- Stopped: terminated, suspended or withdrawn, with the registry's stated reason.
- Verified abstract: a row found by web research is kept only if its number and title were found on a fetched source page, on a second fetched page, or in the verbatim excerpt saved from the source. Everything else is dropped. Abstract text is never stored.
- Lead investigators: overall officials (PI, study chair) with credentials stripped, ranked by Ph2+ trials.
Roadmap
- NowLive registry, PubMed, Open Targets and Crossref data on every view; verified ESMO and ASH abstracts from Parallel web research; fifteen indications including six childhood cancers, glioblastoma, glioma and myeloma, plus a paediatric view. No mock or sample data anywhere.
- NextComplete abstract books: full Crossref harvests for ASCO, AACR and SITC, and session-by-session Parallel pulls for ESMO and ASH, with topline results as structured fields.
- NextReadouts and deals from company disclosures: which trial read out, what was reported, with citations, joined to NCT IDs.
- NextPrograms (asset × indication × segment) so phase advances are tracked per program, backed by a Postgres store.
- LaterWeekly brief per indication; alerts on phase advances and primary completions; a target 'herd' map of pipeline crowding against genetic support.
Motivation: Curing All Disease, Ahead of Schedule.