Use the original ChemrytIQ molecule or a selected analog from generated SMILES.
Prerequisite: Load The Molecule From ChemrytIQ
Before opening ChemrytIQ-PCP, search the molecule in ChemrytIQ by SMILES, InChI, molecule name, or CAS number. Confirm the correct molecule on the ChemrytIQ page, then open the required Chemryt app from that same molecule context so the selected structure is loaded into the app automatically.
What It Does
ChemrytIQ-PCP keeps pharmacophore matching code, target-family profile logic, GNN-backed profile endpoints, feature maps, bioisosteric comparison, metabolic soft spots, and analog-selection tools inside the ChemrytIQ workflow.
Review 2D/3D pharmacophore features, compatibility profile, reference overlays, and activity-cliff context.
Uses server-side profile endpoint and GNN validation summaries where available.
Quick Tutorial
- Open PCP from a ChemrytIQ molecule or select an analog from DeNovo/ToxPred candidate output.
- Confirm the molecule identifiers, SMILES, SELFIES, and current structure preview.
- Run the PCP profile and review pharmacophore feature detection, model metrics, and target profile cards.
- Inspect 2D/3D pharmacophore maps, compatibility profiles, reference examples, and optional overlay layers.
- Use bioisosteric pharmacophore matching or analog optimization panels to compare candidate changes.
- Send promising protected candidates or optimized leads back into PCP or downstream modules for another pass.
Main Areas
| Area | What to review | When to use it |
|---|---|---|
| Molecule identifiers | SMILES, SELFIES, original/analog source, and current structure. | Use before running a profile. |
| Profile output | Target-family matches, GNN summary, pharmacophore compatibility, and confidence details. | Use to interpret target fit. |
| Maps and analogs | 2D/3D maps, ghost features, reference examples, metabolic soft spots, and bioisostere comparison. | Use for medicinal chemistry review. |
ML Model / Computation Used
| Model or method | What it predicts | Implementation details |
|---|---|---|
| PCP message-passing GNN | Target-family profile and activity/pActivity compatibility signals. | PyTorch message-passing GNN trained on BindingDB target-family data; model candidates include pcp_gnn_e70 with seven target families and best-epoch selection by activity ROC-AUC minus pActivity MAE. |
| PCP target-fit baseline | Lightweight target-family/activity baseline for smoke or fallback ranking. | Uses sklearn hashing-SMILES baseline with model.joblib, trained on a smaller BindingDB-derived set for kinase/protease target-family labels. |
Good Practice
Pharmacophore matching is pattern-based support. Confirm target relevance with biological data, structural evidence, and assay context before acting on a PCP score.
Reference Used
This Tutorial page mirrors the ChemrytIQ reference module: ChemrytIQ-PCP.