iDrug

Multi-scenario DTI Prediction and Personalized Drug Recommendation Support Platform. iDrug is an upgraded platform built for multi-scenario drug-target interaction prediction and personalized drug recommendation support. It combines scenario-specific DTI workflows, structured drug and protein knowledge, and patient-aware reranking to provide lightweight, explainable assistance for drug discovery and medication decision support.

Get Started Watch Tutorial

Core Modules

Explore the major workflows of iDrug through personalized recommendation, task-based prediction, structured query, data browsing, and guided learning.

Recommend

Personalized Recommendation

Generate patient-aware drug recommendations by combining DTI-based prediction with age, sex, disease history, allergy history, and insurance-related reranking.

Open Recommend →

Models

Prediction Models

Run multi-scenario DTI workflows for drug repositioning, novel target screening, new drug evaluation, and affinity prediction through a task-oriented prediction workspace.

Open Models →

Explore

Browse drug, protein, and complex information through local-first query workflows, including compound lookup, protein lookup, and 3D complex visualization.

Open Explore →

More Platform Workspaces

Continue from prediction and query workflows into structured data browsing, result management, and guided onboarding.

Data

Data Sources

View the structured local datasets and external reference sources that support prediction, recommendation, metadata matching, and information exploration.

Open Data →

Profile

Review saved results, revisit historical recommendation records, and manage personal prediction outputs through the user profile workspace.

Open Profile →

Tutorial

Follow the guided platform walkthrough with screenshots and step-by-step instructions to quickly understand the main workflows of iDrug.

Open Tutorial →

Platform Architecture

iDrug adopts a three-layer architecture:

Base Layer 1: Multi-scenario DTI Prediction Supports drug repositioning, target screening, drug evaluation, and affinity prediction through task-specific models.
Base Layer 2: Knowledge Query Provides structured drug, protein, and complex information with support for 3D visualization.
Core Layer: Personalized Recommendation Re-ranks candidate drugs with patient-specific features and outputs interpretable results.