Volume 13, Issue 4 (3-2017)                   JSDP 2017, 13(4): 63-78 | Back to browse issues page



DOI: 10.18869/acadpub.jsdp.13.4.63

XML Persian Abstract Print


Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Estimation of protein-coding regions in numerical DNA sequences using Variable Length Window method based on 3-D Z-curve. JSDP. 2017; 13 (4) :63-78
URL: http://jsdp.rcisp.ac.ir/article-1-101-en.html

Abstract:   (428 Views)

In recent years, estimation of protein-coding regions in numerical deoxyribonucleic acid (DNA) sequences using signal processing tools has been a challenging issue in bioinformatics, owing to their 3-base periodicity. Several digital signal processing (DSP) tools have been applied in order to Identify the task and concentrated on assigning numerical values to the symbolic DNA sequence, then applying spectral analysis tools such as the discrete Fourier transform (DFT) to locate the periodicity components. Despite of many advantages of Fourier transform in detection of exotic regions, this approach has some restrictions, such as high computational complexity and disability in locating the small length coding regions. In this paper, we improve the performance of the conventional DFT in estimating the protein coding regions utilizing a Gaussian window with variable length. First, the DNA strands are converted to numerical signals via the 3-D Z-curve method. Z curve is a robust, independent, less redundant approach, and has clear biological interpretation which can be regarded as a useful visualization technique for DNA analysis of any length. In the second stage, non-coding regions besides the background noise components are completely suppressed using the Gaussian variable length window. Also, we use a narrow-band band-pass filter in order to extract the period-3 components with  central frequency. Performance of the proposed algorithm is tested on F56F11.4 from C.elegans chromosome III, also two eukaryotic datasets, HMR195 and BG570,  is compared with other state-of-the-art methods based on the nucleotide evaluation metrics such as sensitivity, specificity, approximation correlation, and precision. Results revealed that, the area under the receiver operating characteristic (ROC) curve is improved from 4% to 40%, in HMR195 and BG570 datasets compared to other methods. Furthermore, the proposed algorithm reduces the number of incorrect nucleotides which are estimated as coding regions.   

Full-Text [PDF 2631 kb]   (149 Downloads)    
Type of Study: Research | Subject: Paper
Received: 2013/06/6 | Accepted: 2016/06/15 | Published: 2017/06/6 | ePublished: 2017/06/6

Add your comments about this article : Your username or Email:
Write the security code in the box

© 2015 All Rights Reserved | Signal and Data Processing

Designed & Developed by : Yektaweb