Fourier transformation for exon prediction

In this bachelor thesis there are described the methods of the prediction of exon. The first part is aimed at the difference between the prokaryotic and eukaryotic organisms, the description of the DNA structure and the explanation of the terms exon and intron. The second section of the theoretic part includes four methods of the prediction of exons namely dynamic programming, neural networks, hidden Markov models and discrete Fourier transform. In the practical part there was created the program called Predikce_exonu that searches exons in nucleotide sequences and works on the principle of the Fourier transform. This algorithm together with 3 freely available programs was tested on 25 sequences and the success of their prediction was described by sensitivity and specificity