Implementation and Development of Signal Processing Tools for Genomic data

Abstract

In literature review we have studied number of algorithms which are used to measure the power spectrum of the DNA sequences with N/3 period property and are helpful in distinguishing the protein coding region and the non-coding region in the DNA sequences. A large number of algorithms have been advanced in last few years for the gene prediction. In my work I have firstly done the analysis on various window functions, digital signal processing algorithm and the mapping scheme. The hanning window function is being used as it provide the best performance for big data sets then we make use of short time Fourier transform as a DSP algorithm which provide better performance with the hanning window. After this analysis I have used another digital signal processing algorithm maximum entropy method which provides much better performance as compared with the short time Fourier transform. The mapping schemes paired- numeric, EIIP is used for numerical conversion of data. The performance depends on the spectral analysis of the windowed data. At each position of the DNA sequence the Fourier transform is calculated of the sequence.