Abstract
This study presents in silico prediction of toxic activities and carcinogenicity, represented by the potential carcinogenicity DSSTox/DBS, based on vector regression with a new Kernel activity, and correlating the predicted toxicity values through a QSAR model, namely: QSTR/QSCarciAR (quantitative structure toxicity relationship/quantitative structure carcinogenicity–activity relationship) described by 2D, 3D descriptors and biological descriptors. The results showed a connection between carcinogenicity (compared to the structure of a compound) and toxicity, as a basis for future studies on this subject, but each prediction is based on structurally similar compounds and the reactivation of the substructures of these compounds.
Acknowledgements
We like to express gratitude for the opportunity to work in the project POSDRU: Invest in people. European Social Fund Operational Program Human Resources Development 2007–2013 Priority Axis 1 “Education and training in support of growth and development of knowledge based society” “Key Area of Intervention 1.5. Support for PhD students and postdoctoral researchers”. Title of the project: Doctoral and Postdoctoral Programs – support for increasing the competitiveness of research in sciences. Contract Code: HRD/159/1.5/S/137750. Gainer: “Alexandru Ioan Cuza” University Iasi and West University Timisoara; in addition, we would like to thank to Prof. Dr. Mircea Mracec for giving access to HyperChem Professional software.
Declaration of interest
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.