chembl-database

Retrieve bioactive molecule, target, and activity data from the ChEMBL database.

2.7k|283|Updated May 13, 2026
One-click install
npx skills add https://github.com/google-deepmind/science-skills --skill chembl-database-google-deepmind
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: chembl-database
Source: https://github.com/google-deepmind/science-skills/tree/main/skills/chembl_database
Command: npx skills add https://github.com/google-deepmind/science-skills --skill chembl-database-google-deepmind

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

It helps you retrieve structured information about bioactive molecules, drug targets, bioactivity (e.g., IC50/Ki), and related chemical structures without manually browsing and reformatting results.

Core Features & Use Cases

  • ChEMBL database querying: Look up molecules/compounds, targets, and assays with consistent JSON outputs.
  • Bioactivity retrieval and normalization: Fetch activities and optionally normalize reported values to nM for fair comparison across studies.
  • Structure-based searches and assets: Run server-side similarity and substructure searches, and download compound images or structure files (SDF/MOL) for downstream analysis.
  • Workflow-ready: Enforces use of the provided utility script for ChEMBL API interactions and requires file-based JSON outputs so an agent can reliably extract IDs and data.

Quick Start

Use the chembl-database skill to fetch EGFR IC50 bioactivities and save the results into /tmp/egfr_ic50.json.

Frequently Asked Questions about chembl-database

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I query ChEMBL bioactivity data and normalize IC50 values to nM?

You can retrieve IC50 and Ki activities from the ChEMBL database and optionally normalize reported bioactivity units to nM for fair comparison across studies using deterministic subcommands. The skill requires file-based JSON outputs via the --output flag for reliable data extraction.

Can I run structure-based similarity and substructure searches on ChEMBL molecules?

Yes, ChEMBL similarity and substructure searches are supported to identify bioactive compounds based on chemical structure. You can query molecules by structural similarity or substructure and retrieve structured JSON outputs containing matching compound identifiers.

How do I download compound images and structure files like SDF or MOL from ChEMBL?

You can download compound images and structure files in SDF or MOL formats from the ChEMBL database for downstream analysis. The skill uses a mandatory utility script for all API interactions to fetch these structural assets and save them locally.

What is the best way to retrieve drug targets and mechanism of action data from ChEMBL?

Retrieving drug targets and mechanism of action data from ChEMBL is handled through dedicated deterministic subcommands. The skill queries the ChEMBL API to fetch structured target and mechanism information, outputting the results to a specified JSON file.

Does the ChEMBL database skill support pagination for large bioactivity query results?

Yes, pagination is supported for ChEMBL database queries to manage large sets of molecule, target, and activity data. The skill's utility script handles API interactions with pagination to ensure complete retrieval of bioactivity records into JSON outputs.

How do I look up specific molecules and compounds in the ChEMBL database for research?

You can look up molecules and compounds in the ChEMBL database using the molecule subcommand to retrieve structured information. The skill enforces JSON output to a specified file so agents can reliably extract compound IDs and bioactivity data.