ADAPTIVE MACHINE LEARNING TECHNIQUE FOR PERIODICITY DETECTION IN BIOLOGICAL SEQUENCES

Researchers devoted considerable effort to detect the periodicity in DNA sequences, namely, the DNA segments that wrap the Histone protein. It is anticipated that periodic dinucleotide signals are indicators of certain important spots ( like binding regions) within a DNA sequence; they are equally spaced along nucleosomal DNA with similar to 10 base-pair period. Positioned nucleosomes are believed to play an important role in transcriptional regulation and for the organization of chromatin in cell nuclei. In this paper, we describe and apply a dynamic periodicity detection algorithm to discover the periodicity of certain dinucleotides in DNA and Protein sequences. Our algorithm is based on suffix tree as the underlying data structure. The proposed approach is suitable to analyze different kinds of data and can serve different targets. It considers the periodicity of alternative substrings like the three dinucleotides AA/TA/TT, in addition to considering dynamic window to detect the periodicity of certain instances of substrings. We tested the applicability, effectiveness and resilience of the proposed approach to noise as compared to the other existing algorithms described in the literature.