Reader Comments

Post a new comment on this article

Referee Comments: Referee 1

Posted by PLOS_ONE_Group on 10 Apr 2008 at 18:51 GMT

Referee 1's Review:

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

The study entitled "Association of common polymorphisms in GLUT9 gene with gout but not with coronary artery disease in a large case-control study" by Klaus Stark et al. aimed to investigate polymorphisms in GLUT9 in both gout and CAD/MI samples. Subjects were ascertained through a large German MI study and both phenotypes were investigated using separate case/control sample sets. The results presented in the study indicate that SNPs in GLUT9 are associated with gout, but not with CAD or MI. This was further enhanced by using WGA data from two large MI studies and identifying no association with CAD in either study. This study is particularly important due to the recent identification of GLUT9 SNPs and association with serum uric acid levels, which has been deemed a risk for CAD. According to the authors, no other studies have examined the genetic effect of GLU9 with CAD or gout.

Overall, the research methods and statistical analyses are adequate for the assessment of the results. However, there are a few issues that should be addressed or discussed in the manuscript. First, it is unclear if there is significant overlap in case/control samples in the gout dataset and with the CAD/MI dataset as both samples are taken from the same larger German MI study. If there is significant overlap, this could skew the results and make for non-independence of the two datasets. This should be addressed and clarified in the results and/or discussion. Second, the authors did not adjust haplotypes for any risk factors, nor was age adjusted for in the CAD phenotype. As age was significantly different between the case/control group, this should be accounted for in the analysis. Additionally, the percentage of gout in the case sample was significantly different that that in the CAD control - this could be a confounding issue if the authors are trying to conclude that GLUT9 SNPs are not associated with CAD independent of gout. Additionally, it was noted that there was a deviation from HWE in the gout cases - was the genotyping further analyzed and scrutinized for genotyping errors, etc as this deviation in HWE could be the cause of the association in the gout samples.

This paper would be most useful to researchers and clinicians studying the causes of gout. However, it should be noted that the entire linkage disequilbrium region of GLUT9 was not taken into account in this paper. It appears from the hapmap data on figure 1 that there are two large LD blocks, of which only one is covered in this manuscript. Thus it seems premature to conclude that SNPs in GLUT9 do not account for MI/CAD in this population, without taking into account all of the genetic variability in the region, especially given the four SNPs genotyped in this study are in relatively high linkage disequilibrium. A better approach, or one that should be employed in future studies, would be to genotype haplotype tagging SNPs in the region and assess them for association. This should be commented on in the discussion.

I commend the authors for using the publicly available WGA data to identify if there is replication with other similar sample sets. The authors may also wish to take advantage of the Framingham SHARE database (http://www.ncbi.nlm.nih.g... ), as there are multiple MI and CAD phenotypes, as well as gout phenotypes. Comparison of the gout phenotypes would be particularly interesting in this paper as the authors are reporting a significant association with this phenotype.

The overall readability of the paper was adequate, although some paragraphs were difficult to comprehend (eg. The last paragraph in the discussion describing association with gout, beginning with "The biological link of GLUT9..."). Sufficient detail, i.e. SNP names, phenotypes, statistical analyses, was presented for others to replicate the results.

Overall, the paper is interesting and warrants the results to be made publicly available. With a few minor edits, I believe this paper worth publishing.