Detecting mtDNA effects with an Extended Pedigree Model: An Analysis of Statistical Power and Estimation Bias

Area of Science:

• Behavioral Genetics
• Mitochondrial Genomics
• Quantitative Genetics

Background:

• Mitochondrial DNA (mtDNA) is vital for cellular functions but its role in human behavior is not well understood.
• Existing research has not fully isolated the specific genetic contributions of mtDNA to behavioral traits.
• Quantifying mtDNA's influence requires sophisticated models that account for nuclear DNA and environmental factors.

Purpose of the Study:

• To develop and validate a novel statistical model for estimating the variance attributable to mitochondrial DNA (mtDNA) in human behavior.
• To determine the sample size and pedigree structure necessary for reliably detecting mtDNA effects on behavior.
• To assess the impact of data imperfections, such as missing kinship data and mtDNA mutations, on model accuracy.

Main Methods:

• Development of a seven-parameter covariance structure model to estimate mtDNA variance in human behavior.
• Utilized extended pedigree data to disentangle mtDNA effects from other genetic and environmental influences.
• Conducted Monte Carlo simulations to evaluate the power of the model across different sample sizes and pedigree structures.

Main Results:

• A sample size of approximately 5,000 individuals is adequate for detecting medium mtDNA effects (mt² = 5%), while 30,000 are needed for small effects (mt² = 1%).
• Deeper pedigrees enhance statistical power to detect mtDNA effects, whereas wider pedigrees decrease it for a fixed sample size.
• Missing kinship data and mtDNA mutations cause underestimation of mtDNA variance and overestimation of nuclear-mtDNA interactions.

Conclusions:

• The proposed covariance structure model effectively quantifies the influence of mitochondrial DNA on human behavior.
• Extended pedigrees are a powerful tool for investigating mtDNA's behavioral impact, with specific structural characteristics optimizing detection power.
• The model demonstrates robustness against common data limitations, showing a low false positive rate and providing reliable estimates.