THESIS
2004
ix, 58 leaves : ill. (some col.) ; 30 cm
Abstract
The research of Gene Predicting Algorithms is a key section in bioinformatics. After brief introduction to bioinformatics, we introduce a new and improved Gene Predicting Algorithm, together with a model to give an analytical explanation of its meaning....[
Read more ]
The research of Gene Predicting Algorithms is a key section in bioinformatics. After brief introduction to bioinformatics, we introduce a new and improved Gene Predicting Algorithm, together with a model to give an analytical explanation of its meaning.
DNA sequences are formed by patches or domains of different nucleotide composition, for which the Jensen Shannon divergence method is often employed to find the boundaries for these compositional domains. By introducing one new parameter α into Jensen Shannon divergence, we find numerically the optimal value to obtain the best accuracy of border finding. We explain this result mathematically, and give the exact expression for this parameter. We then apply this improved Jensen Shannon divergence to artificial sequences and some real DNA sequences. The results demonstrate that this parameter is useful for segmentation of genomic sequences into compositionally homogeneous segments.
Post a Comment