Boosting heritability

Boosting heritability estimates the genetic component of phenotypic variation using high-dimensional linear fixed-effect models combined with multiple sample splitting.

Key Features:

• Multiple sample splitting: Uses a multiple sample splitting strategy to achieve stable and accurate heritability estimates.
• Fixed-effect high-dimensional linear modeling: Employs a fixed-effect model framework for high-dimensional linear models as an alternative to random-effect approaches.
• Integration of recent methodologies: Combines advantageous elements of recent methods to enhance inference in high-dimensional settings.
• Demonstrated on simulated and real data: Validated using both simulated datasets and real-world data.
• Application to Streptococcus pneumoniae antibiotic resistance: Applied specifically to antibiotic resistance data from Streptococcus pneumoniae to assess genetic contributions to resistance.

Scientific Applications:

• Heritability estimation: Quantifies the genetic component of trait variability using high-dimensional fixed-effect models.
• Antibiotic resistance genetics: Investigates genetic contributions to antibiotic resistance in Streptococcus pneumoniae.
• Method benchmarking: Benchmarks inference performance on simulated datasets.

Methodology:

Implements a multiple sample splitting strategy within a fixed-effect high-dimensional linear modeling framework and integrates elements of recent methodological advances.