---
title: "Compound safety profile"
description: Whether a compound is safe enough to advance — off-targets, interactions, and toxicology in one pass.
tags:
  - recipes
  - safety
tools:
  - safety-profile-comprehensive
  - ddi-safety-screen
  - comprehensive-toxicology
generated: true
---

> **Try it in a guest chat**

> Whether a compound is safe enough to advance — off-targets, interactions, and toxicology in one pass. Every prompt below is verbatim from the curated quick-action
> chips, so it runs the real routing. Open one in a guest chat (no account
> needed), edit the compound or target, and send.

**Best fit:** whether a compound is safe enough to advance — off-targets, interactions, and toxicology in one pass.

## Try it — safety profile comprehensive

**Try this prompt** — [open in a guest chat](/chat/new?guest=true&prompt=Run%20comprehensive%20safety%20profiling%20on%20sunitinib%20(fetch%20from%20PubChem)%3A%20screen%20against%20the%20safety%20panel%20for%20off-target%20kinase%20activity%20(hERG%2C%20CYP%2C%20GPCR)%2C%20predict%20drug-drug%20interactions%20with%20common%20co-medications%2C%20and%20classify%20for%20hepatotoxicity%2C%20cardiotoxicity%2C%20and%20teratogenicity%20risk%20-%20do%20the%20predictions%20match%20the%20known%20FAERS%20cardiac%20signals%2C%20and%20recommend%20the%20most%20critical%20safety%20monitoring%20biomarker%20for%20a%20new%20sunitinib%20clinical%20trial%3F).

> Run comprehensive safety profiling on sunitinib (fetch from PubChem): screen against the safety panel for off-target kinase activity (hERG, CYP, GPCR), predict drug-drug interactions with common co-medications, and classify for hepatotoxicity, cardiotoxicity, and teratogenicity risk - do the predictions match the known FAERS cardiac signals, and recommend the most critical safety monitoring biomarker for a new sunitinib clinical trial?

_Seeded from the safety-profile-comprehensive quick-action chip — open it in a guest chat, then change the compound or target._

## Try it — ddi safety screen

**Try this prompt** — [open in a guest chat](/chat/new?guest=true&prompt=Fetch%20methotrexate%20from%20PubChem%2C%20run%20DeepPurpose%20to%20predict%20off-target%20binding%20across%20600%2B%20human%20proteins%2C%20classify%20drug-drug%20interaction%20risk%20with%20common%20co-medications%20(NSAIDs%2C%20trimethoprim%2C%20omeprazole)%2C%20identify%20which%20interactions%20are%20mediated%20by%20renal%20transporter%20competition%20vs%20CYP%20inhibition%2C%20and%20recommend%20the%20co-medication%20contraindication%20with%20the%20highest%20real-world%20clinical%20impact%3F).

> Fetch methotrexate from PubChem, run DeepPurpose to predict off-target binding across 600+ human proteins, classify drug-drug interaction risk with common co-medications (NSAIDs, trimethoprim, omeprazole), identify which interactions are mediated by renal transporter competition vs CYP inhibition, and recommend the co-medication contraindication with the highest real-world clinical impact?

_Seeded from the ddi-safety-screen quick-action chip — open it in a guest chat, then change the compound or target._

## Try it — comprehensive toxicology

**Try this prompt** — [open in a guest chat](/chat/new?guest=true&prompt=Fetch%20bisphenol%20A%2C%20BPS%2C%20and%20BPF%20from%20PubChem%2C%20compute%20RDKit%20descriptors%20(MW%2C%20LogP%2C%20TPSA%2C%20HBD%2FHBA)%2C%20run%20ADMET-AI%20for%20full%20toxicity%20prediction%2C%20dock%20all%20three%20against%20estrogen%20receptor%20alpha%20(PDB%3A%201ERE)%20with%20DiffDock%20to%20compare%20binding%20poses%2C%20and%20run%20DeepPurpose%20to%20predict%20ER-alpha%20affinity%20-%20load%20all%20results%20into%20the%20sandbox%20and%20use%20matplotlib%20to%20generate%20a%20comparative%20toxicology%20dashboard%20(ADMET%20radar%20charts%20%2B%20DiffDock%20pose%20overlays%20%2B%20DeepPurpose%20affinity%20bar%20chart)%2C%20then%20assess%20whether%20BPA%20alternatives%20actually%20have%20cleaner%20ER%20binding%20profiles%3F).

> Fetch bisphenol A, BPS, and BPF from PubChem, compute RDKit descriptors (MW, LogP, TPSA, HBD/HBA), run ADMET-AI for full toxicity prediction, dock all three against estrogen receptor alpha (PDB: 1ERE) with DiffDock to compare binding poses, and run DeepPurpose to predict ER-alpha affinity - load all results into the sandbox and use matplotlib to generate a comparative toxicology dashboard (ADMET radar charts + DiffDock pose overlays + DeepPurpose affinity bar chart), then assess whether BPA alternatives actually have cleaner ER binding profiles?

_Seeded from the comprehensive-toxicology quick-action chip — open it in a guest chat, then change the compound or target._

## Read next

- [Workflows](/docs/workflows) — every workflow and recipe.
- [Which tool?](/docs/which-tool) — match a single goal to a tool.
- [How the orchestrator picks tools](/docs/how-orchestration-works) — the same chain from a plain-language goal.