KeBABS: an R/Bioconductor package for kernel-based analysis of biological sequences
Sprache des Titels:
Englisch
Original Buchtitel:
ASHG 2015 Proceedings
Original Kurzfassung:
The computational analysis of biological sequences is a fundamental task. On the one hand, sequence analysis methods have supplied highly valuable insights into how patterns/motifs in amino acid sequences govern protein structure. On the other hand, a large proportion of our current knowledge about how DNA sequence patterns control transcription factor binding, nucleosome positioning and remodeling, alternative splicing, etc., is the result of computational sequence analysis. In genetics, discriminative sequence analysis is becoming increasingly important to predict potential effects of single-nucleotide variations in the context of surrounding sequences.In the last two decades, kernel methods have been established as an important class of sequence analysis methods. For the classification of sequences, in particular, support vector machines (SVMs) have emerged as a sort of best practice. To apply SVMs for sequence analysis, it is necessary to either use a vectorial representation of the sequence data or to use kernels, that is, positive semi-definite similarity measures for sequences. The use of sequence kernels, however, is not limited to sequence classification. For example, they can also be used for regression tasks and similarity-based clustering.This contribution is devoted to introducing KeBABS, a powerful, flexible, and easy-to-use framework for kernel-based analysis of biological sequences based on the widely used scientific computing platform R. KeBABS is publicly and freely available via the Bioconductor project (for more information, see http://www.bioinf.jku.at/software/kebabs). It includes efficient implementations of the most important sequence kernels, also including variants that allow for taking sequence annotations and positional information into account. For whole abstract see http://www.bioinf.jku.at/publications/2015/ASHG2015_Palme.pdf