Ligand-X: Automated Protein Simulation Platform

Ligand-X is a web application for structural, physical, and chemical analysis of ligand-protein complexes, lowering the barrier to entry for advanced bioinformatics and cheminformatics simulations.

Project Overview

Ligand-X is a comprehensive web-based platform for molecular structure analysis, visualization, and computational chemistry workflows. It serves as a bridge between complex computational chemistry tools and researchers, providing a suite of microservices for cheminformatics and bioinformatics simulation, all accessible through a modern, intuitive web interface. The application supports multiple file formats (PDB, CIF, mmCIF, SDF), integrates with the RCSB PDB database, and features an interactive molecular editor and viewer, powered by Ketcher and Mol* respectively.

Key Features

Ligand-X offers tools for every stage of the computer aided drug discovery pipeline, executed locally on your machine using battle tested methodologies:

Structure Cleaning

Automated preparation of raw PDB structures, including protonation and residue repair.

Theory Demo

Molecular Docking

High-throughput ligand pose prediction and affinity scoring using AutoDock Vina.

Theory Demo

Molecular Dynamics

GPU-accelerated simulations with OpenMM to study complex biological movements.

Theory Demo

Relative Binding Free Energy (RBFE)

Compare binding affinities across a series of ligands using network-based alchemical calculations.

Theory Demo

Absolute Binding Free Energy (ABFE)

Calculate the exact binding affinity of a single ligand to a protein target with rigorous thermodynamic methods.

Theory Demo

Quantum Chemistry

Detailed electronic analysis and ligand parameterization using DFT calculations.

Theory Demo

ADMET Prediction

Predict molecular properties and ADMET characteristics for drug-likeness assessment.

Learn More

Technical Stack

Layer	Technologies
Frontend	React, Next.js, TypeScript, Mol* (3D visualization), Ketcher (molecular editor)
Backend	FastAPI, Celery, Redis, PostgreSQL
Chemistry	RDKit, AutoDock Vina, OpenMM, OpenFF, ORCA, OpenFE, Boltz-2
Deployment	Docker Compose, Conda, GPU support

Architecture

Ligand-X uses a modular microservices architecture where each computational task runs in its own isolated service:

Structure Service: Handles PDB/SDF parsing, component identification, and structure cleaning
Docking Service: Manages molecular docking workflows with AutoDock Vina
MD Service: Orchestrates molecular dynamics simulations with OpenMM
QC Service: Manages quantum chemistry calculations with ORCA
ADMET Service: Predicts molecular properties and drug-likeness
Alignment Service: Performs protein structure alignment
Ketcher Service: Provides interactive molecular editing
Boltz-2 Service: AI-based binding affinity predictions

Each service can be deployed independently, allowing you to use only the tools you need. The gateway coordinates requests across services and manages job queuing through Redis and Celery workers.

Getting Started

Ligand-X is containerized with Docker for easy deployment. Simply clone the repository, from there you can use make build to build the Docker images, and make up to start all services. If you are going to be making changes to the code, you should use make dev to start all services in development mode, with frontend hot reload enabled.

The application will be available at http://localhost:3000
with the API gateway at http://localhost:8000.
the rabbitmq management interface will be available at http://localhost:15672.
the celery workers will be available at http://localhost:5555/workers/.

authentication can be setup using environment variables.

For detailed setup instructions, refer to the Ligand-X documentation on the github page:

Other Projects

Konstantin N 2026-02-25 PROJECTS
React Next.js PDBFixer MolStar Ketcher OpenFF OpenMM OpenMMForceFields OpenFE AutoDock Vina Quantum Chemistry