Archived Comments for:
A genome-wide association study using international breeding-evaluation data identifies major loci affecting production traits and stature in the Brown Swiss cattle breed
Hossein Jorjani, Swedish University of Agricultural Sciences
12 June 2014
We have been informed by one of the sources providing data that a genotyping mistake (at the SNP clustering level) has been discovered that potentially has affected the genotypes of 2254 animals (32%) used in our study (Guo et al., BMC Genetics 2012, 13:82 doi:10.1186/1471-2156-13-82).
The results of our preliminary re-analysis of the data, after the exclusion of the animals with possible genotyping mistake, show the following. The total number of QTLs that reached the highest significance level per chromosome for the 9 reported traits has changed from 60 to 63. Out of the 63 QTLs detected in the re-analysis, 31 map to the same SNPs as before. The rest of them (32 QTLs) map to new SNPs on the same chromosomes, or to other chromosomes. The change in the average p-value is negligible (from 2.67E-07 to 2.63E-07), even though individual p-values may change considerably. The beta values for the strongest signal (the QTLs with the highest significance level) per chromosome are stable; however the beta values for the other QTLs on the same chromosome show much variation.
Our assessment of the preliminary re-analysis of data, after exclusion of the animals with possible genotyping mistake, is that our general conclusions in the published manuscript still hold. The general conclusions were 1) international genetic evaluation data is a useful resource for identifying genetic factors influencing complex traits in livestock; 2) several genome wide significant association signals could be identified in the Brown Swiss population; 3) there are several associations and plausible candidate genes that deserve further exploration (e.g. molecular dissection to explore the potential economic impact and the genetic mechanisms underlying these production traits in cattle). However, there are many small details in the published manuscript that need to be re-visited and reported again, and are not precise as they are now.
We propose to repeat the study, after the genotyping mistakes have been corrected (given we are granted the permission to use the new data), and submit a new manuscript. Given the results of our preliminary re-analysis of the data mentioned above, we believe the repetition of the study would benefit from including comparison of within/across country GWAS results to detect possible genotype-environment interaction.
Detection of a genotyping mistake
12 June 2014
We have been informed by one of the sources providing data that a genotyping mistake (at the SNP clustering level) has been discovered that potentially has affected the genotypes of 2254 animals (32%) used in our study (Guo et al., BMC Genetics 2012, 13:82 doi:10.1186/1471-2156-13-82).
The results of our preliminary re-analysis of the data, after the exclusion of the animals with possible genotyping mistake, show the following. The total number of QTLs that reached the highest significance level per chromosome for the 9 reported traits has changed from 60 to 63. Out of the 63 QTLs detected in the re-analysis, 31 map to the same SNPs as before. The rest of them (32 QTLs) map to new SNPs on the same chromosomes, or to other chromosomes. The change in the average p-value is negligible (from 2.67E-07 to 2.63E-07), even though individual p-values may change considerably. The beta values for the strongest signal (the QTLs with the highest significance level) per chromosome are stable; however the beta values for the other QTLs on the same chromosome show much variation.
Our assessment of the preliminary re-analysis of data, after exclusion of the animals with possible genotyping mistake, is that our general conclusions in the published manuscript still hold. The general conclusions were 1) international genetic evaluation data is a useful resource for identifying genetic factors influencing complex traits in livestock; 2) several genome wide significant association signals could be identified in the Brown Swiss population; 3) there are several associations and plausible candidate genes that deserve further exploration (e.g. molecular dissection to explore the potential economic impact and the genetic mechanisms underlying these production traits in cattle). However, there are many small details in the published manuscript that need to be re-visited and reported again, and are not precise as they are now.
We propose to repeat the study, after the genotyping mistakes have been corrected (given we are granted the permission to use the new data), and submit a new manuscript. Given the results of our preliminary re-analysis of the data mentioned above, we believe the repetition of the study would benefit from including comparison of within/across country GWAS results to detect possible genotype-environment interaction.
Respectfully,
Hossein Jorjani
Competing interests
None declared