Chemoinformatics in Multi-target Drug Discovery for Anti-cancer Therapy: In Silico Design of Potent and Versatile Anti-brain Tumor Agents

Alejandro      Speck-Planche; Valeria      V. Kleandrova; Feng      Luan; M.      Natalia D. S. Cordeiro

doi:10.2174/187152012800617722

Abstract

A brain tumor (BT) constitutes a neoplasm located in the brain or the central spinal canal. The number of new diagnosed cases with BT increases with the pass of the time. Understanding the biology of BT is essential for the development of novel therapeutic strategies, in order to prevent or deal with this disease. An active area for the search of new anti-BT therapies is the use of Chemoinformatics and/or Bioinformatics toward the design of new and potent anti-BT agents. The principal limitation of all these approaches is that they consider small series of structurally related compounds and/or the studies are realized for only one target like protein. The present work is an effort to overcome this problem. We introduce here the first Chemoinformatics multi-target approach for the in silico design and prediction of anti-BT agents against several cell lines. Here, a fragment-based QSAR model was developed. The model correctly classified 89.63% and 90.93% of active and inactive compounds respectively, in training series. The validation of the model was carried out by using prediction series which showed 88.00% of correct classification for active and 88.59% for inactive compounds. Some fragments were extracted from the molecules and their contributions to anti-BT activity were calculated. Several fragments were identified as potential substructural features responsible of anti-BT activity and new molecular entities designed from fragments with positive contributions were suggested as possible anti-BT agents.

Keywords: Anti-BT activity, Atom-centered fragments, Brain tumor, BT cell lines, Chemoinformatics, Drug–receptor interactions, Fragment, Functional group counts, Linear additivity, Linear discriminant analysis, Multi-target QSAR, Quantitative contributions, ROC curve, Spectral moments, Substructural alerts