Discovery of evolutionary patterns in ribosomal RNA data using markov models


The paper deals with the construction of the phylogenetic tree andmultiple sequence alignment of ribosomal RNA datasetsdownloaded from Silva comprehensive rRNA database usingmarkov models. The new method is based on the concept ofcomparing the similarity/dissimilarity between two Markov modelsusing Kullback–Leibler divergence for construction of pair-wisedistance matrix. The alignment accuracy (sum of pair’s scores) ofthe multiple sequence alignment of these datasets is compared withClustalX software (progressive alignment). The statisticalsignificance and the deviation of the alignment accuracy fromClustalX software was found by analysis of variance (ANOVA).Our results using the new method showed good agreement with theClustalX-Multiple Sequence Alignment and phylogenetic tree andconfirming the use of probabilistic Markov models in rRNA datasetanalysis. All datasets and computer codes written in MATLAB areavailable upon request from the first author([email protected]).