ProLIF: a library to encode molecular interactions as fingerprints

Abstract

Interaction fingerprints are vector representations that summarize the three-dimensional nature of interactions in molecular complexes, typically formed between a protein and a ligand. This kind of encoding has found many applications in drug-discovery projects, from structure-based virtual-screening to machine-learning. Here, we present ProLIF, a Python library designed to generate interaction fingerprints for molecular complexes extracted from molecular dynamics trajectories, experimental structures, and docking simulations. It can handle complexes formed of any combination of ligand, protein, DNA, or RNA molecules. The available interaction types can be fully reparametrized or extended by user-defined ones. Several tutorials that cover typical use-case scenarios are available, and the documentation is accompanied with code snippets showcasing the integration with other data-analysis libraries for a more seamless user-experience. The library can be freely installed from our GitHub repository ( https://github.com/chemosim-lab/ProLIF ).

Keywords: Docking; Interaction fingerprint; Molecular dynamics; Python; Structural biology; Virtual screening.

Fig. 1

Ligand interaction network for the ergotamine agonist bound to the 5-HT1B receptor. Each interaction is shown as a dashed line between the residue and the ligand, and the width of the line is linked to the frequency of the interaction in the simulation. Only interactions occurring in at least 30% of frames are shown here

Fig. 2

Tanimoto similarity matrix of ligand–protein interactions between each frame of the MD trajectory

Fig. 3

Residue interaction network for the bovine rhodopsin. Residues are colored by transmembrane domain (TM). Interactions that only appear in the active (PDB 6FK6) or inactive (PDB 1U19) state of the receptor are respectively shown in green or orange, and the ones that appear in both are in grey. Each residue node is scaled based on its number of interactions. For clarity, interactions that occur within the same TM (as labelled by GPCRdb) and interactions between residues that are less than 3 residues apart are not shown, as well as hydrophobic interactions (as defined in the implementation) and residues that did not participate in any interaction

Fig. 4

Interaction network between the β2 adrenoceptor (ADRB2) and G protein complex (Gαs and Gβ1). ADRB2 residues are shown as rectangles in shades of green, and G protein residues are shown as ellipses in shades of blue for Gαs and in yellow for Gβ1. For ADRB2, ICL denotes the intracellular loops while TM corresponds to the transmembrane domains. For Gαs, the common Gα numbering (CGN) system is used [37]. Each node is scaled by its number of interactions. Inter and intra protein interactions are respectively shown as plain and dashed lines. Residues that do not participate in GPCR-G protein interactions are not shown, and interactions between covalently bonded residues or residues of the same helix (as labelled by GPCRdb) are hidden