This app derives from a notebook (Model Building Demo) in the ReinventCommunity repo. The aim of this app is to generate a QSAR model for a target protein that you are interested. QSAR modeling is an estimation of the activity/property/toxicity of new chemical entities in the drug discovery process. In the demo, SMILES that known activities on the DRD2 receptor were used. SMILES are transformed into fingerprints and act as the input in the code. Since REINVENT only supports pickled scikit-learn models, the output files are in (.pkl) format. The output file either can be used by (Reinvent: Reinforcement Learning Module) or (Reinvent: Scoring Module) apps. For more information, please see the tutorial page of Reinvent Apps.

Example Use Case: Generating QSAR models for target protein

Limitations:
- Only scikit-learn models (.pkl) are accepted by REINVENT
- Since the demo notebook contains only classification methodology for the random forest, the app provides this option currently.

Technology: Random Forest, scikit-learn, Quantitative structure-activity relationship

Metrics: AUC score metrics from demo data are represented here. Also, you can check the Example Output of this app.