April 2010
Volume 51, Issue 13
ARVO Annual Meeting Abstract  |   April 2010
Copy Number Variants Associated With Refractive Error
Author Affiliations & Notes
  • Y.-J. Li
    Biostatistics and Bioinformatics,
    Duke University Medical Center, Durham, North Carolina
  • T. Young
    Duke University Medical Center, Durham, North Carolina
  • Footnotes
    Commercial Relationships  Y.-J. Li, None; T. Young, None.
  • Footnotes
    Support  EY, Research to Prevent Blindness
Investigative Ophthalmology & Visual Science April 2010, Vol.51, 2480. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Y.-J. Li, T. Young; Copy Number Variants Associated With Refractive Error. Invest. Ophthalmol. Vis. Sci. 2010;51(13):2480.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Purpose: : Segmental copy-number variations (CNVs) in the human genome affect phenotypic diversity- contributing both to normal variation and to increased disease susceptibility. To identify CNV regions associated with refractive error, we conducted a genome-wide CNV analysis of genotypes from 3024 DNA samples from 708 twin sets and relatives.

Methods: : 3024 genomic DNA samples from a twin cohort were genotyped with the Illumina human610-quad beadchip, consisting of 550,000 tag SNPs and 50,000 common region CNV markers. The PennCNV software program was used to identify CNV regions based on hidden Markov modeling. We determined whether each SNP was present or absent in any of the sample’s CNV events. A linear regression model analysis was performed in the R package with mean sphere (Sph) and mean spherical equivalent (SE) as outcome variables. To fit model assumptions of normality, both Sph and SE were transformed using the ENQT method. The predictor variable of the model was the presence or absence of a CNV event for a particular marker. To ensure adequate sample size, linear regression models were run only for markers that were included in CNVs for at least 0.4% of the data (12 samples).

Results: : Of 600,470 SNPs in the analysis, 295,507 (49.2%) were involved in at least one CNV event, and 88,606 (14.8%) were involved in 12. 2957/ 3024 (97.8%) sampled individuals had available phenotypic data. Using a threshold significance level of 0.005, significant SNPs clustered in 7 CNV chromosomal regions: 1) 3p11.1 (p=0.0005, 0.000221 for Sph and SE, respectively); 2) 3q13.31 (p=0.000708, 0.000515); 3) 7q11.23 (p=0.001894, 0.000767; 4) 9p13.1 (p=0.0004, 0.001217); 5) 10p11.1 (p=0.0001, 0.0000973; 6) 10p11.21 (p=0.000986, 0.001202); and 7) 13q12.12 (p=0.001077, 0.00033). The 10p11.1 and 10p11.21 regions were associated with a high myopic refractive error- the average Sph (SE) in samples with the CNV event was -4.06 D (-4.52 D) and -8.69 D (-9.19 D), respectively, compared to +0.06 D (-0.29 D) in the entire dataset. The identified CNV regions at chromosomes 3q13.31, 7q11.23, 9p13.1, 13q12.12 were also associated with myopia, although the average Sph (SE) values were more moderate than for the chromosome 10 regions. The average Sph (SE) for these regions were -1.66 D (-1.91 D), -0.82 D (-1.23 D), -0.86 D (-1.09 D), and -0.93 D (-1.24 D), respectively. The 3p11.1 CNV region was associated with hyperopia with an average Sph (SE) of +1.17 D (+0.84 D).

Conclusions: : CNV events at specific chromosomal intervals have been identified, and show significant association to various refractive error phenotypes in a large-scale, high-resolution genotyping study.

Keywords: refractive error development • genetics • myopia 

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.