Title: Stochastic entropy QSAR for the in silico discovery of anticancer compounds: prediction, synthesis, and in vitro assay of new purine carbanucleosides
Authors: González-Díaz, Humberto ×
Viña, Dolores
Santana, Lourdes
De Clercq, Erik
Uriarte, Eugenio #
Issue Date: Feb-2006
Series Title: Bioorganic & medicinal chemistry vol:14 issue:4 pages:1095-107
Abstract: A Markov model based QSAR is introduced for the rational selection of anticancer compounds. The model discriminates 90.3% of 226 structurally heterogeneous anticancer/non-anticancer compounds in training series. External validation series were used to validate the model; the 91.8% containing 85 compounds, not considered to fit the model, were correctly classified. The model developed is afterwards used in a simulation of a virtual search for anticancer compounds never considered either in training or in predicting series. The 87.7% of the 213 anticancer compounds used in this simulated search were correctly classified. The model also shows high values for specificity (0.89), sensitivity (0.91), and Mathews correlation coefficient (0.79). In addition, the present model compares better-to-similar with respect to other four models elsewhere reported if one takes into consideration 26 comparison parameters. Finally, we exemplify the use of the model in practice with the design of a new series of carbanucleosides. The compounds evaluated with the model were synthesized and experimentally assayed for their antitumor effects on the proliferation of murine leukemia cells (L1210/0) and human T-lymphocyte cells (CEM/0 and Molt4/C8). The more interesting activity was detected for the compound 5a with a predicted probability of 80.2% and IC(50) = 27.0, 27.2, and 29.4 microM, respectively, against the above-mentioned cellular lines. These values are comparable to those for the control compound Ara-A.
ISSN: 0968-0896
Publication status: published
KU Leuven publication type: IT
Appears in Collections:Laboratory of Virology and Chemotherapy (Rega Institute)
× corresponding author
# (joint) last author

Files in This Item:

There are no files associated with this item.

Request a copy


All items in Lirias are protected by copyright, with all rights reserved.

© Web of science