Purchase this article with an account.
Nina Riddell, Loretta Giummarra, Sheila Gillard Crewther; Systematic integrated comparison of omics studies highlights similarities in the biological basis of animal and human myopia. Invest. Ophthalmol. Vis. Sci. 2016;57(12):3611.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The relevance of animal models to understanding human myopia was questioned by a recent review that identified only 4 similarities between loci in the CREAM genome-wide association study (GWAS) and differentially-expressed genes from two animal microarray studies of refractive error. We expanded this approach to systematically compare exploratory animal transcriptome and proteome studies of environmentally-driven refractive error with all GWAS of myopia indexed in the NHGRI catalog.
Inclusion criteria required that studies were exploratory investigations of refractive error, and that animal studies compared optically-induced ocular growth changes to normal growth controls. A search of the NHGRI catalog and PubMed identified 14 GWAS, 7 transcriptome, and 9 proteomics studies meeting these criteria. For GWAS, genes associated with each SNP were identified based on author-reports. For transcriptome and proteomics studies, differentially-expressed genes/proteins were identified using the original author’s statistical criteria. The resulting lists were converted to a common identifier, and compared at the single gene and pathway enrichment level using Lists2Networks and EnrichR.
Comparing within methods, only 7% of transcriptome findings and 22% of protein findings have been replicated more than once. Highly replicated results (genes/proteins implicated ≥3 times) did not show specificity for the direction of ocular growth, with the exception of UNC80 in transcriptome studies, and SERPINF1, APOA1, COL1A1 and COL1A2 in proteome studies. Comparing across methods, 25 genes were implicated in both transcriptome and GWAS, and the overlap between genes associated with myopia-SNPs in humans and genes down-regulated transcriptionally in animal myopia was statistically significant (p<.001 Bonferroni corrected). There were no single gene commonalities between proteome and GWAS, however further enrichment analyses revealed that genes implicated by all three methods affect common signaling pathways.
These findings support the continued use of animal models for investigations of refractive error, and imply that environmental and genetic mechanisms may converge on the same pathways to control eye growth across species.
This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.
This PDF is available to Subscribers Only