Modeling for Evaluation of Significant Features in siRNA Design

Abstract

AbstractRNAi is the most conserved phenomenon occurring in eukaryotes, where it controls the developmental process through gene regulation. Recently, exogenously generated siRNA mediated RNAi has drawn greater significance in functional genomics and therapeutic applications like cancer, HIV and neurodegenerative diseases specially in mammalian system. Computational designing of efficient sequence specific siRNAs against the gene of interest deploy many guidelines, which are based upon sequence to thermodynamic features as a pivotal determinants of effective siRNA sequences, but identification of optimal features needed for efficient designing are yet to be deciphered in the assurance of better efficacy. Till date many computational tools are available, but no tool provide the accurate gene specific siRNA sequences with absolute efficacy therefore study of suitable features of siRNA design is very smoldering issue to be solved in the present scenario. In the present work, we have applied ant colony optimization technique to indentify the features of siRNA up to considerable amount of accuracy and further the results are modeled using four independent models such as linear regression, ANCOVA, libSVM and liblinear with the conclusion that linear features are preferentially superior then thermodynamic features while both group of features are important in the efficacy prediction of siRNA. The results are highly coherence with prior investigations and highlight the importance of sequential features in effective siRNA design.KeywordsSupport Vector MachineFeature SelectionHuman Immunodeficiency Virus TypeProtein Function PredictionsiRNA DesignThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.