The Analytical Problem
Chemotypic analysis entails unique problems and potentials
related to the analysis of chemotypic indices as opposed to the more
traditional chemical descriptors used in substructure, QSAR and
similarity analysis. Each chemotypic index is the informational
equivalent of a high-dimensional array of chemical descriptors, and,
although standard chemometric packages are capable of generating and
analyzing a high-dimensional array of chemical descriptors, they are
unprepared for handling the massive informational content entailed in
multiple chemotypic indices and for the unique questions addressed
when using chemotypic indices. For example, the simple problem of computing the
number of unique ring systems present in a chemical library creates
programming needs not easily handled by most chemometricians using
standard software. Grouping those ring systems by shared ring-system
skeletons introduces a hierarchical complexity of a still higher
order. The Meqi software was developed to address this need in chemotypic
analysis to work with multiple hierarchically-related chemotypic indices.
The Meqi Software
The Meqi software runs in a Windows PC environment as a command-line
executable that operates on SD files of any size. It reads a CSV command file
for its instructions. This command file is generated from a user-modifiable
fill-in-the-blank programming template. This template, as illustrated, is an
extensively annotated Excel spreadsheet. By adding and modifying sections
and subsections of the template, the user can generate
endless combinations of chemotypic indices reflecting the defining atom and
bond properties, many 2D global and local positioning considerations,
and various user-defined transformations of atom and bond labels.
After executing the instructions in the command file, a flexible menu of
command-line table operations is provided for examining and summarizing the
resulting table of indices in a manner particularly suited to chemotypic
analysis.
The Meqi software is distributed under a freeware license and can be downloaded upon agreeing to the terms of the license.
The Meqi Code
The Meqi source code was
written using Microsoft Visual C++ Version 6.0 but
employs only standard ASCII functionalities. It was developed primarily as
a research tool for flexibly exploring methodologies and indices
appropos chemotypic analysis. As such, it provides an interface for
flexibly specifying a wide variety of chemotypic indices and is
structured so as to facilitate enhancements for any one familiar with
C++. It is being made available for commercial and non-commercial use
without charge so as to encourage enhancements to the code itself or
to related proprietary or nonproprietary software that lie outside this
developer's interests but which are clearly needed for broader use of
the code and related chemotypic methodologies.
The Meqi code can be downloaded, examined and used free of charge upon only upon agreeing to the source code licensing agreement.
Last updated March 23, 2009