Aarhus University Seal / Aarhus Universitets segl

Comparing power and precision of within-breed and multibreed genome-wide association studies of production traits using whole-genome sequence data for 5 French and Danish dairy cattle breeds

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review


  • Irene van den Berg, Universite Paris-Saclay
  • ,
  • Didier Boichard, Universite Paris-Saclay
  • ,
  • Mogens Sandø Lund

The objective of this study was to compare mapping precision and power of within-breed and multibreed genome-wide association studies (GWAS) and to compare the results obtained by the multibreed GWAS with 3 meta-analysis methods. The multibreed GWAS was expected to improve mapping precision compared with a within-breed GWAS because linkage disequilibrium is conserved over shorter distances across breeds than within breeds. The multibreed GWAS was also expected to increase detection power for quantitative trait loci (QTL) segregating across breeds. GWAS were performed for production traits in dairy cattle, using imputed full genome sequences of 16,031 bulls, originating from 6 French and Danish dairy cattle populations. Our results show that a multibreed GWAS can be a valuable tool for the detection and fine mapping of quantitative trait loci. The number of QTL detected with the multibreed GWAS was larger than the number detected by the within-breed GWAS, indicating an increase in power, especially when the 2 Holstein populations were combined. The largest number of QTL was detected when all populations were combined. The analysis combining all breeds was, however, dominated by Holstein, and QTL segregating in other breeds but not in Holstein were sometimes overshadowed by larger QTL segregating in Holstein. Therefore, the GWAS combining all breeds except Holstein was useful to detect such peaks. Combining all breeds except Holstein resulted in smaller QTL intervals on average, but this outcome was not the case when the Holstein populations were included in the analysis. Although no decrease in the average QTL size was observed, mapping precision did improve for several QTL. Out of 3 different multibreed meta-analysis methods, the weighted z-scores model resulted in the most similar results to the full multibreed GWAS and can be useful as an alternative to a full multibreed GWAS. Differences between the multibreed GWAS and the meta-analyses were larger when different breeds were combined than when the 2 Holstein populations were combined.

Original languageEnglish
JournalJournal of Dairy Science
Pages (from-to)8932-8945
Number of pages14
Publication statusPublished - 1 Nov 2016

    Research areas

  • genome-wide association studies (GWAS), meta-analysis, multibreed, whole genome sequence

See relations at Aarhus University Citationformats

ID: 107903246