Title:

A modest proposal on Virtual Screening: We hold these truths to be
self-evident

Author:
Glen Kellogg
Virginia Commonwealth University

Abstract:

When in the Course of biomolecular binding events, it becomes necessary
for one species to displace the molecules of solvent which have occupied
the site of another, and to assume a low-energy bound conformation in this
site, a decent respect to the Laws of Chemistry and Physics requires
enumeration of the Forces which impel them to associate.  We hold these
truths to be self-evident, that all biomolecular binding is concerted,
that Binding Free Energy is endowed by the Creator with certain
unalienable Properties, that among these are Hydrogen Bonds, Coulombic
Forces, London Forces and the pursuit of Entropy.  --That to calculate
these properties, Scoring Functions have been created, deriving their
usefulness from correlations with experiment, --That whenever any Form of
Scoring Function becomes inaccurate to these ends, it is the Right of the
User to alter or to dismiss it, and to institute a new Scoring Function,
laying its foundation on such principles of and organizing its algorithms
in such form, as to him seem most likely to effect his Predictions and Job
Security.  Prudence, indeed will dictate that Scoring Functions long
established should not be changed for light and transient causes; and
accordingly all experience hath shewn, that Users are more disposed to
suffer, while inaccuracies are sufferable, than to right the predictions
by abolishing the functional forms to which they are accustomed.  But when
a long train of false positives and false negatives, pursuing invariably
the same Lead Compounds evinces a design to reduce the productive hits in
Assays, it is their right, it is their duty, to throw off such Scoring
Functions, and to obtain new Computational Tools for their future
security.  The results from conventional Virtual Screening are results of
repeated missed opportunities and failures, all having in direct object
the establishment of an unsupported paradigm for Drug Discovery.  To prove
this, let Facts be submitted to a candid world.

Virtual Screening does not, in usual cases, include the effects of solvent
Molecules within the active Site that are in practice supporting the
binding.

Virtual Screening has affected to ignore the Ionization States of
Functional Groups from both the Protein and Ligand; if accounted for,
these effects may be quite substantial.

Virtual Screening has deemed it reasonable to calculate Hydrophobic
Interactions with Algorithms both overly simple and unsubstantiated.

Virtual Screening provides no redress for the phenomenon of Entropy, that
which most assuredly is a vital and inalienable component of Binding Free
Energy.

We, therefore, submit a modest and CPU-efficient proposal to reform and
improve the performance of Virtual Screening with empirical Scoring
Functions derived from the experimental measurement of Water-Octanol
solvent partitioning.  We, also, put forward parallel methods in which
Calculations of Electrostatic Potential as performed by Zap and other
Computer programs can be utilized to a similar result.

OpenEyeCupVI_Kellogg.pdf