A method to address differential bias in genotyping in large-scale association studies
2007

Addressing Bias in Genotyping for Disease Studies

Sample size: 7500 publication 10 minutes Evidence: moderate

Author Information

Author(s): Vincent Plagnol, Jason D. Cooper, John A. Todd, David G. Clayton

Primary Institution: University of Cambridge

Hypothesis

Can methodological improvements reduce differential bias in genotyping in large-scale association studies?

Conclusion

The study successfully implemented methodological improvements that reduced the false-positive rate in genotyping by addressing differential bias.

Supporting Evidence

  • The adapted algorithm increased the number of confidently scored SNPs from 5,294 to 7,446.
  • Over-dispersion decreased from 17% to 7.5% with the new methodology.
  • The use of fuzzy calls helped to avoid bias from treating uncertain calls as missing.
  • Statistical tests showed significant improvements in data quality with the adapted algorithm.

Takeaway

This study found better ways to check DNA samples for diseases, which helps scientists get more accurate results.

Methodology

The study adapted genotyping algorithms to minimize biases from different DNA sourcing and used 'fuzzy' calls to handle uncertain genotypes.

Potential Biases

Differential bias in genotype calling due to different DNA sourcing can inflate false-positive rates.

Limitations

The study acknowledges that some over-dispersion remains and that the inclusion of non-Caucasian samples increased over-dispersion.

Participant Demographics

The study involved 3,750 type 1 diabetes cases and 3,480 controls from the 1958 British Birth Cohort.

Statistical Information

P-Value

0.036

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1371/journal.pgen.0030074

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication