ProLIF

ProLIF generates interaction fingerprints as a Python library by encoding three-dimensional interaction patterns of molecular complexes (ligands interacting with proteins, DNA, or RNA) into vector representations to support structure-based analyses in drug discovery and computational biology such as virtual screening and machine learning.

Key Features:

• Versatile input compatibility: Supports complexes derived from molecular dynamics trajectories, experimental structures, and docking simulations for consistent fingerprint generation across diverse structural inputs.
• Vectorized 3D interaction fingerprints: Produces vector representations that encode three-dimensional interaction patterns within a complex for use in quantitative analyses.
• Comprehensive and extensible interaction definitions: Represents interactions across combinations of ligands, proteins, DNA, and RNA, with the ability to reparameterize existing interaction types and add user-defined interactions.

Scientific Applications:

• Structure-based virtual screening: Uses interaction fingerprints as compact descriptors to prioritize ligands based on binding interaction patterns.
• Similarity analysis and predictive modeling: Enables similarity comparisons and input features for machine learning and predictive modeling of binding behavior.
• Target assessment and ligand optimization: Supports assessment of binding interactions for target characterization and iterative ligand optimization.
• Analysis of simulation and docking outputs: Facilitates examination of complexes produced by molecular dynamics simulations and docking workflows.

Methodology:

Generates vector interaction fingerprints from molecular complexes using definable interaction types, accepts inputs from molecular dynamics trajectories, experimental structures, and docking simulations, and allows reparameterization and addition of user-defined interaction definitions.