Protein-Ligand Interaction Fingerprints

Protein-ligand interaction fingerprints can be used to determine which amino acid residues are crucial to docking interactions. Docking can be used as an initial screen. Protein ligand interaction is more specific in that it tells you what bonds are used. Protein ligand interaction fingerprints can be used to find compounds similar to known binders or to existing drugs. It can be useful for drug repurposing.

Protein ligand interaction profile (plip) is a module available for python. It can give you hydrophobic interactions, hydrogen bonds, pi stacking, pi cation interaction etc.

https://github.com/pharmai/plip

Protein–ligand interactions are coded into binary vectors. Each vector consists of seven bits for each ligand binding site residue, which represents different types of interactions.

Given two plips you can calculate the similarity in binding between two protein ligand complexes. The fingerprints can then be paired with the autodock energy to build a machine learning model that predicts the energy of protein ligand complexes with the same ligand and protein with new poses.

Letś take the structure of covid19 in complex with some inhibitors.

Using autodock, generate a bunch of poses.