If these findings can be validated by additional studies, they will have a significant impact on the analysis and interpretation of GWAS data. GWAS commonly identify genes that are—in terms of their function—at first sight hard to link to the phenotype studied. The connection of numerous DNAse hypersensitive sites harboring GWAS SNPs with promoters of distant genes—as suggested by this study—offers a plausible explanation for these observed associations. In addition, the work suggests that seemingly unconnected variants have common transcription factor networks. If both findings turn out to be true, they would largely help us to understand and disentangle the specific disease studied, since the authors show how apparently unrelated genes can be linked to disease.
These findings also help us understand the links among different diseases, because they provide a framework for elucidating whether diseases share a common transcription factor network and thus may be mechanistically related.
Going forward, association studies could be performed differently than done at present, namely by examining hits only on regulatory DNA. This would increase the statistical power (and thus the ability to detect disease-associated variants), as it would decrease the statistical correction needed for the multiple testing required for genomewide studies.
Comments
Erasmus MC
If these findings can be validated by additional studies, they will have a significant impact on the analysis and interpretation of GWAS data. GWAS commonly identify genes that are—in terms of their function—at first sight hard to link to the phenotype studied. The connection of numerous DNAse hypersensitive sites harboring GWAS SNPs with promoters of distant genes—as suggested by this study—offers a plausible explanation for these observed associations. In addition, the work suggests that seemingly unconnected variants have common transcription factor networks. If both findings turn out to be true, they would largely help us to understand and disentangle the specific disease studied, since the authors show how apparently unrelated genes can be linked to disease.
These findings also help us understand the links among different diseases, because they provide a framework for elucidating whether diseases share a common transcription factor network and thus may be mechanistically related.
Going forward, association studies could be performed differently than done at present, namely by examining hits only on regulatory DNA. This would increase the statistical power (and thus the ability to detect disease-associated variants), as it would decrease the statistical correction needed for the multiple testing required for genomewide studies.
View all comments by Christiane ReitzMake a Comment
To make a comment you must login or register.