ElectroShape: fast molecular similarity calculations incorporating shape, chirality and electrostatics

Abstract

We present ElectroShape, a novel ligand-based virtual screening method, that combines shape and electrostatic information into a single, unified framework. Building on the ultra-fast shape recognition (USR) approach for fast non-superpositional shape-based virtual screening, it extends the method by representing partial charge information as a fourth dimension. It also incorporates the chiral shape recognition (CSR) method, which distinguishes enantiomers. It has been validated using release 2 of the Directory of useful decoys (DUD), and shows a near doubling in enrichment ratio at 1% over USR and CSR, and improvements as measured by Receiver Operating Characteristic curves. These improvements persisted even after taking into account the chemotype redundancy in the sets of active ligands in DUD. During the course of its development, ElectroShape revealed a difference in the charge allocation of the DUD ligand and decoy sets, leading to several new versions of DUD being generated as a result. ElectroShape provides a significant addition to the family of ultra-fast ligand-based virtual screening methods, and its higher-dimensional shape recognition approach has great potential for extension and generalisation.

Improving the accuracy of ultrafast ligand-based screening: incorporating lipophilicity into ElectroShape as an extra dimension

“…The partial charges were re-assigned using the modified MMFF94 [8] charges implemented in CCG MOE [6] (the difference between standard MMFF94 and MOE's MMFF94x is that in the later, hydrogen atoms on alkane carbons have a non-zero partial charge). These were the partial charges that resulted in the highest enrichment factor at 1% in the previous studies [1].…”

“…Following our previous work, the effect on performance of inclusion of atomic lipophilicity values will be studied in enrichment experiments that will compare use of partial charges as a fourth dimension (as in [1], atomic lipophilicity as a fourth dimension, and their combination into a five dimension method.…”

“…The implementation of ElectroShape (calculation of the centroids, and then of the corresponding descriptors) is described in detail in Armstrong et al [1]. Briefly, given a collection of atom positions, the procedure defines the creation of four or five distinct ''centroids''.…”

“…In a previous work [1], we introduced ElectroShape, an ultra-fast, descriptor-based tool for ligand-based virtual screening. This built on the original ultra-fast shape recognition (USR) method [2,3] by incorporating the ability to distinguish chiral molecules [4] and adding atomic electrostatic information (via atomic partial charge) as a fourth dimension.…”

How reliable are ligand-centric methods for Target Fishing?

“…This similarity can be in terms of 2D chemical structure (Nettles et al, 2006) using circular fingerprints (Rogers and Hahn, 2010), 3D molecular properties (Cortés-Cabrera et al, 2013) using Ultrafast Shape Recognition variants (Ballester and Richards, 2007;Armstrong et al, 2010;Ballester, 2011) or NCI-60 bioactivity spectra (Holbeck et al, 2010) using cellular fingerprints (Füllbeck et al, 2009). Note that there are similarity-based methods that are not ligand-centric.…”