Pick a cell line: filter or search below to narrow 39,665 human cancer lines and organoids. Each card shows a depth chip: how well described the line is compared with others of the same cancer type (hover the i for details). Conflict flags warn you where sources disagree about a line.
Think like a clinical trials manager: build a cohort to test in the lab
- How many lines does your point need? A single line is one tumour in a dish. In a clinical trial we recruit a cohort because patients vary; an in vitro cohort needs enough members to carry the variation you plan to talk about. Four lines is already a panel; eight to ten starts to look like a cohort; beyond that, think about what each additional member adds.
- Generalisable or niche? Decide before you build: are the conclusions meant to travel to a broad patient population, or to a specific sub-cohort (a subtype, a treatment state, an ancestry)? Let that choice drive the filters above, and check the clinical-applicability notes on each member rather than assuming a famous line is representative.
- What variation will you accept? Set your tolerance: which characteristics must every member share (the ✓ rows in the comparison table), and which are you happy to see differ (the ✗ rows)? Differing rows are information, not noise - they are the heterogeneity your experiment can interrogate.
- Play. Build a panel, remove a member that does not earn its place, rebuild, and watch how the shared and differing rows move. The tool is designed for iteration; the comparison table below updates with every change.
A panel of authenticated, well-matched lines chosen for the question is usually the least clinically irrelevant option available; no line is a patient.