Skip to main content
ARS Home » Research » Publications at this Location » Publication #264932

Title: A high-resolution cattle CNV map by population-scale genome sequencing

item Bickhart, Derek
item HOU, YALI - University Of Maryland
item Schroeder, Steven - Steve
item ALKAN, CAN - Howard Hughes Medical Institute
item CARDONE, MARIA - University Of Bari
item MATUKUMALLI, LAKSHMI - George Mason University
item SCHNABEL, ROBERT - University Of Missouri
item VENTURA, MARIO - University Of Bari
item TAYLOR, JEREMY - University Of Missouri
item EICHLER, EVAN - Howard Hughes Medical Institute
item Sonstegard, Tad
item Van Tassell, Curtis - Curt
item Liu, Ge - George

Publication Type: Abstract Only
Publication Acceptance Date: 2/22/2011
Publication Date: 5/10/2011
Citation: Bickhart, D.M., Hou, Y., Schroeder, S.G., Alkan, C., Cardone, M.F., Matukumalli, L.K., Schnabel, R.D., Ventura, M., Taylor, J.F., Eichler, E.E., Sonstegard, T.S., Van Tassell, C.P., Liu, G. 2011. A high-resolution cattle CNV map by population-scale genome sequencing. CONFERENCE ON THE BIOLOGY OF GENOMES [abstract]. p. 32.

Interpretive Summary:

Technical Abstract: Copy Number Variations (CNVs) are common genomic structural variations that have been linked to human diseases and phenotypic traits. Prior studies in cattle have produced low-resolution CNV maps. We constructed a draft, high-resolution map of cattle CNVs based on whole genome sequencing data from 70 individuals by integrating complementary structural variant (SV) discovery approaches (CGH and high density SNP arrays) and experimental validations (PCR and FISH). Our sequence data comprised a total of ~90× coverage of the reference genome (~ 270 Gb) distributed among 5 taurine and 3 indicine breeds. To identify CNVs from these sequence data, we adopted and fine tuned two well-established computational approaches: read depth (RD; using WSSD) and read pair (RP; using VariationHunter) analyses. Our initial RD results identified 2,789 duplications/insertions and 156 deletions while the RP results revealed multiple deletions. Discovered CNVs had an average length of 20 kb (RD) and represented ~70 megabases of the cattle reference genome. We integrated these new SVs with the CNV call sets made from CGH array and SNP array based methods. To confirm our combined CNV dataset, we created new CGH arrays and performed PCR assays. Our results confirmed previously published CNV regions and validated the new candidate CNVs. These CNVs span multiple annotated cattle genes, which are significantly enriched for specific biological functions such as immunity, lactation, reproduction and rumination. In summary, the individually constructed cattle CNV maps provide a valuable resource for future cattle genomic research.