Editor-in-Chief İlkay Küçükgüzel Associate Editor Esra Tatar Online ISSN 2630-6344 Publisher Marmara University Frequency Quarterly (January-April-July-October) Abbreviation J.Res.Pharm. Former Name Marmara Pharmaceutical Journal
Journal of Research in Pharmacy 2019 , Vol 23 , Issue 2
Modeling to predict the cytotoxicity of SiO2 and TiO2 nanoparticles
1Pharmaceutical Sciences Research Center, School of Pharmacy, Kermanshah University of Medical Sciences, Kermanshah, Iran
2Pharmaceutical Analysis Research Center and Faculty of Pharmacy, Tabriz University of Medical Sciences, Tabriz, Iran
DOI : 10.12991/jrp.2019.133 The objective of the current study was to design a suitable model to predict the cytotoxicity induced by SiO2 and TiO2 nanoparticles in different conditions using computational models. To achieve this, we employed various statistical approaches such as linear regression, as well as artificial neural networks and support vector machine (non-linear models). The effective input parameters of the SiO2 nanoparticles were particle size, particle concentration, and cell exposure time. In the case of the TiO2 nanoparticles, the particle size and concentration served as input variables. Cell viability was considered the output response for both nanoparticles. The modeling was performed using both linear and non-linear methods. In addition, an external validation analysis was conducted to evaluate the predictability of the models by splitting the data into training and test data. The best models to predict cell viability were the models developed by artificial neural network. The results of this investigation indicate that non-linear models could be superior to linear models in predicting cell viability for SiO2 and TiO2 nanoparticles. Keywords : Artificial neural network; cytotoxicity; modeling; nanoparticles
Marmara University