TarKG

Query drug-target relationships from a local TarKG TSV dataset.

130|4|Updated Mar 13, 2026

Search Tags:#retrieval #knowledge-graph #pharmacology #local-file #pathway #drug-target #tarkg

Ready to run

npx skills add https://github.com/QSong-github/DrugClaw --skill tarkg

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: TarKG
Source: https://github.com/QSong-github/DrugClaw/tree/main/skills/dti/tarkg
Command: npx skills add https://github.com/QSong-github/DrugClaw --skill tarkg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

TarKG enables querying a canonical drug-target knowledge graph to reveal relationships between drugs and their protein targets, with contextual pathways and disease associations to support analysis and interpretation.

Core Features & Use Cases

Retrieval of drug-target interactions with relation labels, diseases, and pathway context.
Quick lookups for drugs and targets within TarKG for rapid hypothesis generation.
Local, lightweight reasoning over a structured TarKG TSV dataset to support pharmacology insights.

Quick Start

Configure the TarKGSkill with the TSV path and run a query for a drug to discover its targets and context.

Frequently Asked Questions about TarKG

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

FAQPage Schema

How do I query drug-target relationships from a local TSV file for pharmacology analysis?▼

To query drug-target relationships, configure the retrieval interface with your local TSV path to extract and explore drug, target, and relation triplets. This enables lightweight pharmacology reasoning and hypothesis generation directly from your structured dataset.

What format does my drug-target knowledge graph dataset need to be for local retrieval?▼

Your drug-target knowledge graph dataset must be a local TSV or CSV file containing columns for drug, target, and relation, with optional disease and pathway columns. This structured format supports pathway-informed interpretation and mechanism analysis.

Can I find disease and pathway context alongside drug-target interactions in a single lookup?▼

Yes, you can find disease and pathway context alongside drug-target interactions in a single lookup. The retrieval interface returns relation labels, associated diseases, and contextual pathways to support safety assessment and mechanism analysis.

What is the best way to generate quick hypotheses for drug discovery using a local knowledge graph?▼

The best way to generate quick drug discovery hypotheses is to run targeted queries for specific drugs or protein targets against your local knowledge graph. This retrieval process rapidly maps contextual pathways and disease associations for interpretation.

Do I need any external dependencies or APIs to explore drug-target triplets?▼

No external dependencies or APIs are required to explore drug-target triplets. The skill performs local, lightweight reasoning over a structured TarKG TSV dataset loaded directly from your configured file path.

When should I use a local TSV triplet file instead of a database for drug-target retrieval?▼

You should use a local TSV triplet file for drug-target retrieval when you need lightweight, immediate reasoning without external database setup. It is ideal for rapid lookups, mechanism analysis, and pathway-informed interpretation in isolated environments.