How to estimate the measurement error variance associated with ancestry proportion estimates

Area of Science:

• Population Genetics
• Statistical Genetics
• Genomic Research

Background:

• Accurate individual ancestry proportion estimation is crucial for genomic research.
• Measurement error (ME) in these estimates can inflate type I errors and reduce statistical power.
• Existing methods for estimating ME variance have limitations, particularly when dealing with multiple ancestral populations (k>2).

Purpose of the Study:

• To propose and evaluate a novel method for estimating the variance of measurement error (ME) in individual ancestry proportion estimates.
• To compare the performance of the proposed internal consistency measures with the traditional repeated measurement (RM) approach for ME variance estimation.
• To investigate the impact of the number and size of marker subsets on the accuracy of ME variance estimation.

Main Methods:

• Extending existing internal consistency measures to estimate ME variance.
• Comparing internal consistency estimates with ME variance estimated via the repeated measurement (RM) approach.
• Utilizing simulated data with varying numbers of ancestral populations (k) and marker subset configurations.

Main Results:

• Internal consistency measures provided less biased and more precise ME variance estimates than the RM approach, irrespective of k.
• Both methods showed improved performance with a larger number of similarly sized marker subsets.
• The proposed method effectively addresses ME in ancestry proportion estimates.

Conclusions:

• Internal consistency measures offer a superior approach for estimating ME variance in individual ancestry proportions compared to the RM method.
• Optimizing marker subsetting strategies enhances the accuracy of ME variance estimation.
• Implementing ME correction methods based on these findings will improve the reliability of association tests and other genomic analyses.