Synthetic evidence record A
Example record text for interface review. This is not a publication, DOI, PMID, or retrieved finding.
Explore a QSAR study workflow
Move through a fixed example of how QSARify connects a research question, evidence review, dataset preparation, model review, and prediction.
From question to prediction preview
Step 1 of 8
Swipe or scroll to see all 8 steps
Example input
Research question
The first step starts with a question, then keeps the evidence trail attached as the study moves forward.
Which fictional target has enough example records to demonstrate a QSAR study workflow?
This invented prompt stays in the browser and is never submitted.
Snapshot attached
Evidence review
A report can keep sources beside the target statements they support. These cards show the shape of that evidence record.
Synthetic
Synthetic database record B
Illustrates provenance fields that can accompany an evidence item.
Illustrative labels
Target assessment
Target identity, evidence role, modality, and activity coverage are separate fields. Every value below is synthetic.
Fictional target
Not a real target
Example kinase A
DEMO-TARGET-01 · fictional identifier
Biological relevanceIllustrative
Evidence qualityIllustrative
QSAR readinessIllustrative
The identifiers and scores are illustrative; no database lookup was performed.
Fixture snapshot
Dataset review
This small table shows how a dataset preview can expose endpoint and unit fields before model training.
| Fixture compound | Endpoint | Unit | Activity label | Source |
|---|---|---|---|---|
| DEMO-CMPD-001 | IC50 | µM | Example active | Fixture |
| DEMO-CMPD-002 | IC50 | µM | Example inactive | Fixture |
| DEMO-CMPD-003 | IC50 | µM | Example active | Fixture |
| DEMO-CMPD-004 | IC50 | µM | Example inactive | Fixture |
These rows are bundled with the page and remain unchanged.
No training
Training setup preview
A real study requires explicit endpoint, split, feature, and model choices. This public fixture page does not start training.
- Endpoint
- IC50 · synthetic fixture
- Data split
- Preview only
- Feature preparation
- Not run
- Model job
- Not created
Review preview
Candidate review
A trained candidate would be reviewed before deployment. There is no model, validation run, or performance metric in this demo.
Model candidate
Preview onlyNot created
There is no training run, metric, or downloadable artifact in this fixture.
No inference
Prediction interface
A playground accepts a compound and returns a model result. This demo only shows a sample state, with no model execution.
Fictional inputDEMO-CMPD-001OutputPreview unavailable
No model or experimental result is associated with this example.
No deployment
Deployment preview
QSARify deployments require a reviewed model and an explicit owner decision. The reviewer demo cannot publish or modify a model.
No deployment exists
No URL, model artifact, or prediction endpoint is created by this fixture.