A brief comparison of DSP and HMM methods for gene finding

Authors:
Sajid A. Marhon;Stefan C. Kremer
Affiliations:
University of Guelph, Guelph, ON, Canada;University of Guelph, Guelph, ON, Canada
Venue:
Proceedings of the ACM Conference on Bioinformatics, Computational Biology and Biomedicine
Year:
2012

Citing 3
Cited 0

A Generalized Hidden Markov Model for the Recognition of Human Genes in DNA

Proceedings of the Fourth International Conference on Intelligent Systems for Molecular Biology
Two Methods for Improving Performance of a HMM and their Application for Gene Finding

Proceedings of the 5th International Conference on Intelligent Systems for Molecular Biology
Identification of Protein Coding Regions Using the Modified Gabor-Wavelet Transform

IEEE/ACM Transactions on Computational Biology and Bioinformatics (TCBB)

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we briefly compare the performance of gene finding methods which are based on hidden Markov models (HMMs) and digital signal processing (DSP). We apply the methods to three benchmark datasets consisting of sequences from various species (mammalian, vertebrate and non-vertebrate) to investigate the strength and weakness of these two classes of methods from different aspects. We study the effect of training on the HMM-based methods. In addition, we analyze the effect of the threshold and window length parameters on the performance of the DSP-based methods. In our work, we present the receiver operating characteristic (ROC) plots of the DSP-based methods and the numerical results of applying the HMM and DSP-based methods to the three datasets. In addition, the plots of the prediction accuracy of the DSP-based methods when they are applied to exons of specific lengths and with different values of the window length parameter are presented in this study.