R-GC-test

R-GC-test implements a penalized mixture logistic regression framework to model and test maternal–fetal and intra-fetal genetic conflicts that contribute to disease risk during human pregnancy.

Key Features:

• Penalized mixture logistic regression: Uses a mixture model-based penalized logistic regression to model heterogeneous genetic effects.
• Unified modeling of two conflict types: Integrates analyses of conflicts between maternal and fetal genes and conflicts within the fetal genome due to opposing maternal and paternal effects.
• Missing paternal genotype handling: Accommodates scenarios with absent paternal genetic information common in family-based studies.
• Variable selection procedure: Implements a variable selection approach to identify significant genetic features contributing to disease risk.
• Simulation-based evaluation: Demonstrates power and false-positive control across simulated sample sizes and allele frequencies.

Scientific Applications:

• Genetic conflict testing: Testing hypotheses about genetic conflicts between maternal and fetal genomes in pregnancy-related traits.
• Intra-fetal effect detection: Detecting opposing effects of maternal and paternal alleles within the fetal genome.
• Association studies of pregnancy outcomes: Association testing for pregnancy-related diseases such as small for gestational age (SGA).
• Family-based studies with missing parents: Analysis of family-based genetic data when paternal genotypes are missing.

Methodology:

Implements a mixture model-based penalized logistic regression with a variable selection procedure, accommodates missing paternal genotypes, and is evaluated using simulation studies.