Development of a Protein-Ligand Extended Connectivity (PLEC) Fingerprint and Its Application for Binding Affinity Predictions.

Abstract

<div>Fingerprints (FPs) are the most common small molecule representation in cheminformatics. There are a wide variety of fingerprints, and the Extended Connectivity Fingerprint (ECFP) is one of the best-suited for general applications. Despite the overall FP abundance, only a few FPs represent the 3D structure of the molecule, and hardly any encode protein-ligand interactions. Here, we present a Protein-Ligand Extended Connectivity (PLEC) fingerprint that implicitly encodes protein-ligand interactions by pairing the ECFP environments from the ligand and the protein. PLEC fingerprints were used to construct different machine learning (ML) models tailored for predicting protein-ligand affinities (pK<sub>i/d</sub>). Even the simplest linear model built on the PLEC fingerprint achieved R<sub>p</sub>=0.83 on the PDBbind v2016 "core set”, demonstrating its descriptive power. The PLEC fingerprint has been implemented in the Open Drug Discovery Toolkit (https://github.com/oddt/oddt).</div>

Multi-reference poly-conformational computational methods for de-novo design, optimization, and repositioning of pharmaceutical compounds

“…The proposed computational algorithm extends the currently available methods [20][21][22][23] and introduces additional search flexibility via the use of the compound conformers. The proposal is to compare multiple possible shapes, adopted via varying environmental conditions, of the same molecule (i.e., conformers) rather than just a single shape that was used before.…”

“…where compounds (Chloroquine [31], Remdesivir [32], and Favipiravir [33]) have recently demonstrated significant efficacy against SARS-COV-2, whereas JQ1, Apicidin, and Haloperidol are already marketed compounds well-known for their efficacy against other disease indications. One hundred conformers for each of the reference molecules were generated at the MMFF94 level of theory [34] and each conformer was ODDT-fingerprinted [20] and saved in the MongoDB database [35].…”

“…The ODDT implementation [20] of ElectroShape fingerprints [22] has been selected to demonstrate the proposed approach because these fingerprints are considered to be state-of-the-art in ligand-based virtual screening experiments [36,37], and they are not limited to binary values. 4.…”

“…The alternative overlaying is performed by comparing shape-based descriptors (a.k.a conformer-level 3D fingerprints). An example of such an approach is ElectroShape implemented in the ODDT package [20] and is based on the algorithm that incorporates shape, chirality, and electrostatics [21,22], and represents each conformer via a fixed-length vector of real-valued numbers. Similarly the E3FP package [23] also utilizes an alignment-invariant 3D representation of molecular conformers as a fixed-length binary vector for each conformer.…”

Selecting Machine-Learning Scoring Functions for Structure-Based Virtual Screening

“…ODDT handles ligand molecules in various file formats, generates features for protein-ligand complexes (e.g. BINANA[51], PLEC[59]) as well as ligands (e.g. USR[60], USRCAT[61]) and builds SFs using established ML algorithms such as RF or NN.…”