Poster Presenter
Anticancer Activity Modeling Using Molecular Descriptors
Delvin Nodarse, Enrique Molina, Yovani Marrero
and J.M. García de la Vega
Spain
Cancer has become a mankind nightmare since it can be caused
by multiple factors with a huge variety of origins. All known treatments
involve very undesirable side effects. For this reason, researchers
try their efforts in the search for new chemical entities presenting
antitumor activity. Attempting to overcome this problem it has been
investigated the use of ligand-based classification models for the
rational selection/identification or design/optimisation of new lead
anticancer from virtual database.
Predictive quantitative structure-activity relationship
(QSAR) models of anticancer compounds were obtained by means of linear
discriminant analysis using topologic (2D) as well as a generic algorithms
for the selection of the best subset of variables. Quantitative models
found for describing the anticancer activity are significant from
a statistical point of view. Molecular descriptors included in our
QSAR models allow the structural interpretation of the biological
process, evidencing the main role of the molecular structure.
In this communication, several quantitative models for the discrimination
of antitumorals have been obtained using the TOpological MOlecular
COMputer Design strategy (TOMOCOMD approach) [1,2]. A variable data
conformed by ca. 600 compounds reported actives against any kind of
cancer has been presented as a helpful tool not only for theoretical
chemist but also for other researchers in this area. The validated
models classify correctly a high percent of compound in both training
and external prediction data sets. They showed high Matthews' correlation
coefficient. The TOMOCOMD-CARDD (Computer-Aided Rational Drug Design)
approach implemented in this work was successfully compared with reported
models for anticancer selection. It is expected that these useful
quantitative structure-activity relationship equations can be used
in the identification of previously unknown antitumorals compounds.
[1] Y. Marrero et al., Int.
J Mol. Sci., 5, 276 (2004).
[2] A. Meneses-Marcel et al. J. Biomol. Screen. 13, 785 (2008).
|
