Skip to main content
ARS Home » Research » Publications at this Location » Publication #231861

Title: A genome-wide analysis of array-based comparative genomic hybridization (CGH) data to detect intra-species variations and evolutionary relationships.

Author
item MITRA, APRATIM - UNIV OF MARYLAND
item Liu, Ge - George
item SONG, JIUZHOU - UNIV OF MARYLAND

Submitted to: PLOS ONE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/13/2009
Publication Date: 11/24/2009
Citation: Mitra, A., Liu, G., Song, J. A genome-wide analysis of array-based comparative genomic hybridization (CGH)data to detect intra-species variations and evolutionary relationships. PLoS One. Nov. 24; 4(11):e7978.

Interpretive Summary: Array Comparative Genomic Hybridization (CGH) can reveal chromosomal aberrations in the genomic DNA. These amplifications and deletions at the DNA level are important for animal health and production. While a large number of approaches have been proposed, the noises related to the hybridization are one of the main obstacles to interpret array CGH results and improve the detection resolution. In this study, we implemented and tested a novel genome-wide method using wavelet power spectrum to remove the noises. We analyzed data from 32 bovine individuals from 5 breeds. Our approach correctly classified 4 out of the 5 breeds. Comparisons of the clustering results showed that it has superior detection sensitivity in most of the cases. Our results suggest that our method can also be used for uncovering potential evolutionary relationships between closely related individuals.

Technical Abstract: Array-based comparative genomics hybridization (CGH) has gained prevalence as a technique of choice for the detection of structural variations in the genome. In this study, we propose a novel genome-wide method of classification using CGH data, in order to reveal putative phylogenetic relationships and pedigree information. We analyzed data from 32 bovine individuals from 5 breeds. The data was first denoised before we performed feature extraction with the Haar wavelet. The wavelet power spectrum was then calculated and the spectrum profiles were classified using Ward’s hierarchical clustering. Our approach correctly classified 4 out of the 5 breeds. Pair-wise comparisons of the clusters using the exact F-test showed that they were significantly different (p<0.05) in all except two cases. CGH data can be used to detect structural variations in the genome, and our results suggest that our method can also be used for uncovering potential phylogenetic relationships between closely related individuals.