Estimating genetic parameters of survival distributions: a multifactorial model

Area of Science:

• Biostatistics
• Quantitative Genetics
• Epidemiology

Background:

• Age of onset (AO) in genetically influenced illnesses often correlates between relatives.
• AO can reflect an individual's inherited liability to disease.
• Existing survival models typically do not account for these familial correlations in AO or liability.

Purpose of the Study:

• To develop and evaluate a survival model based on the gamma distribution that incorporates genetic variability in AO.
• To investigate the relationship between correlations in AO and correlations in underlying genetic liability.
• To assess the accuracy and precision of parameter estimation using maximum likelihood estimation (MLE) for this model.

Main Methods:

• Utilized the gamma distribution to model age of onset (AO) distributions.
• Incorporated genetic liability by allowing the hazard frequency parameter of the gamma distribution to be a function of inherited liability.
• Conducted simulations with varying gamma model orders and analyzed data from simulated twin pairs (monozygotic and dizygotic) using MLE.

Main Results:

• Simulations demonstrated that correlations in AO between relatives can substantially differ from correlations in genetic liability.
• MLE provided reasonably accurate parameter estimates for the gamma models.
• Parameter estimation precision decreased as the order of the gamma process increased.

Conclusions:

• The proposed gamma distribution model effectively accounts for genetic variability in age of onset.
• The distinction between AO correlations and liability correlations is crucial for accurate genetic inference.
• The order of the gamma process impacts the precision of parameter estimates, influencing inferences about age of onset in genetic studies.