> For the complete documentation index, see [llms.txt](https://docs.neurolabai.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.neurolabai.io/examples-and-use-cases/known-compound-investigation.md).

# Investigate a Known Compound

This walkthrough uses **Donepezil** to show the most mature Neurolab application path: resolve a compound already represented in the governed knowledge graph, run the three-member known-compound ensemble, and inspect one structured report across coordinated views.

The example is a frozen historical beta run, not medical advice or a claim that every returned value is experimentally validated.

## Start with a narrow research question

Instead of asking whether Donepezil is “good” or “effective,” ask:

> What mechanism did this release rank first, what returned records are relevant to that interpretation, and what reliability warnings limit the claim?

Keep **Basic** mode selected and enter the compound name.

![Donepezil entered in the Basic compound input before submission](https://3908477117-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fgze1K0yofnhFNmM9nEyt%2Fuploads%2Fgit-blob-307f01d39a08da765cc806beed284520eb7c3b5d%2F01-query.jpg?alt=media)

## Confirm the report, then read reliability

The frozen response returned:

| Field               | Returned value                                | How to read it                                                |
| ------------------- | --------------------------------------------- | ------------------------------------------------------------- |
| MoA rank 1          | `acetylcholinesterase inhibitor`, score 1.062 | A within-head ranking, not a probability                      |
| MoA seed agreement  | 3 of 3, high                                  | Ensemble consistency, not proof of correctness                |
| Cognition rank 1    | `attention`, score 0.775                      | A ranked cognition label                                      |
| Cognition agreement | 2 of 3, medium                                | A reason to keep the cognition result as a shortlist          |
| MoA coverage        | 8 binding/MoA-relevant KG edges; threshold 21 | `low_kg_coverage`; do not let the top score hide this warning |
| Brain map           | receptor-decoder; 6 informative receptor hits | Exploratory and explicitly not decision-grade                 |

The report summary makes the top mechanism easy to see, but the reliability block determines how strongly it may be described.

![Donepezil report interpretation workspace showing mechanism cognition and reliability-aware context](https://3908477117-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fgze1K0yofnhFNmM9nEyt%2Fuploads%2Fgit-blob-e5706b3bf52f38c1d676251cf08e0751dcd77555%2F04-chat-report%20\(2\).jpg?alt=media)

## Use each workspace for a different question

1. **Inference:** Did the request finalize, and which compound did the release resolve?
2. **Raw Payload:** What exact keys, arrays, nulls, warnings, and release metadata were returned?
3. **Chat Report:** Can the application explain one narrow report question without adding unsupported claims?
4. **Knowledge Graph:** Which report-local nodes, edges, paths, evidence records, and sources were returned?
5. **Brain Explorer:** Which channel is active, and what method and limitation labels govern it?

In this run, the report-local graph contained 94 nodes, 106 edges, and 13 source records. Those counts describe the returned support artifact; they do not mean that 106 independent studies validate the top mechanism.

![Donepezil report-local evidence universe with scope filters and artifact counts](https://3908477117-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fgze1K0yofnhFNmM9nEyt%2Fuploads%2Fgit-blob-89d89238f81381a3868387e3fbbfb04c3d154144%2F05-knowledge-graph.jpg?alt=media)

## Read the brain view literally

The captured default is **Legacy specific map (diagnostic)** and is explicitly labeled unvalidated and diagnostic-only. That is the interpretation boundary.

The visual can help a researcher inspect relative parcel and network patterns produced by this release. It must not be described as measured activation, a subject scan, an fMRI result, or proof that the compound causes activity in a displayed region.

![Donepezil Brain Explorer with the diagnostic channel and visualization controls visible](https://3908477117-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fgze1K0yofnhFNmM9nEyt%2Fuploads%2Fgit-blob-22089be93458483e411589f37f3d5169a345180e%2F06-brain-explorer.jpg?alt=media)

## Save a reproducible artifact

Export the Raw Payload and retain the input, route, model version, KG release, warnings, and generation time. The example in this guide is a dated, sanitized capture; use the release identities in your own export when reproducing an investigation.

For the next step, use the same report to [trace one mechanism claim](/examples-and-use-cases/trace-a-report-claim.md) from rank to records, support scope, source identity, and limitation.

***

**Example status:** captured August 6, 2026 with model `v43.262-deleak-edgefeat-transformerconv-B0-3seed`. Current navigation can differ while the frozen values retain their historical interpretation.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.neurolabai.io/examples-and-use-cases/known-compound-investigation.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.