Desenvolvimento e validação de método de análise de sequências genômicas baseada em padrões de entropia, coeficiente de clusterização e periodicidade

Abstract

Genomic sequences carry a wide range of information on organism that compose it. Obviously, by reason that great similarity of this information and functions, it is expected that each sequence can belong to many organisms with a similar probability. However, each genome carries within itself certain peculiarities that can be extracted using appropriate tools. In this context , this paper proposes a methodology for the analysis of genomic sequences of bacteria , using some measures that are particularly important : The entropy of triples ( Sn ) , the quantification of frequency 3 (P3) in a sequence , the clustering coefficient ( D ) and the percentage of GC . The method proposed here allows us to infer which a particular organism genome sequence may belong, being feasible for use in Metagenomics. The results of this study demonstrate the effectiveness of this method, 100 % of the organisms were identified in the samples studied (VP). On the other hand, a large number of bodies which did not belong samples were found (FP), which indicates the high similarity of certain sequences, corroborating some studies indicate that the genome carries ortholog sequences, common to countless organisms.