AMIDE v2: High-Throughput Screening Based on AutoDock-GPU and Improved Workflow Leading to Better Performance and...

Area of Science:

• Computational chemistry
• Drug discovery
• Bioinformatics

Background:

• Molecular docking is crucial for drug discovery and target identification but often demands extensive computational resources.
• Existing methods can be slow and require high-performance computing (HPC) solutions, limiting accessibility and throughput.
• The AutoMated Inverse Docking Engine (AMIDE) was initially developed for HPC-based inverse docking.

Purpose of the Study:

• To enhance the speed and efficiency of the AMIDE inverse docking tool.
• To improve the performance, usability, and compatibility of AMIDE for both personal computers (PCs) and HPC environments.
• To enable high-throughput inverse screening capabilities.

Main Methods:

• AMIDE version 2 was re-engineered using AutoDock-GPU for significant speed improvements.
• The programming workflow underwent a complete revision, focusing on performance, bug fixes, and parallelization.
• Compatibility was expanded for both PC and HPC systems.

Main Results:

• AMIDE version 2 achieves up to 12.4 times faster docking speeds compared to version 1 for 100 runs.
• Reverse docking of a ligand across 87 proteins is completed in 23 minutes on a single Graphics Processing Unit (GPU).
• An exponential increase in computation speed was observed with the increasing number of GPUs used.

Conclusions:

• AMIDE version 2 represents a substantial advancement in computational drug discovery, offering significant speed-ups.
• The tool is now optimized for high-throughput inverse screening, making complex computations more accessible.
• The use of GPUs and improved workflow drastically reduces the time required for large-scale inverse docking tasks.