| name | Compound Triage |
| description | Narrow a hit list to the compounds worth a person's bench time, and say why the rest were dropped. |
| category | Discovery |
Compound Triage
A screen produces more hits than anybody can follow. The value of this pass is
in what it removes and in the reasons it records for removing them.
When to use
- A screen or virtual screen returns a hit list.
- A series has to be narrowed before it goes to the bench.
Steps
- Read
experiments for what has been tried. Re-testing a compound already
found to be an assay artefact is the commonest waste in discovery and it is
entirely preventable.
- Remove the known artefacts first — aggregators, reactive groups,
assay-interference chemotypes. Removed with the reason, not silently.
- Check the properties that decide developability, not just potency. A
potent compound with no viable route to exposure is a paper, not a candidate.
- Look for series, not singletons. A structure-activity relationship across
several related compounds is evidence; a single potent hit is a lottery
ticket.
- State what would confirm each survivor, and what a human would have to do
at the bench to get it.
- Record the drops. Every removed compound gets a reason on
experiments
or the program row, because the next screen will surface it again.
Output
A short ranked list with what would confirm each one, and an explicit record of
what was dropped and why. Nothing here asserts activity in a person; that needs
wet results and the people qualified to read them.
