A general and efficient method for estimating continuous IBD functions for use in genome scans for QTL

Area of Science:

• Genetics and Genomics
• Bioinformatics
• Computational Biology

Background:

• Identity by descent (IBD) matrix estimation is crucial for Quantitative Trait Loci (QTL) mapping using variance component models.
• Current methods estimate IBD discretely, limiting computational efficiency and memory usage in genome-wide scans.
• A novel approach is proposed to estimate IBD as continuous functions.

Purpose of the Study:

• To develop and evaluate a new method for estimating IBD as continuous functions.
• To improve the computational efficiency and memory requirements of genome scanning for QTL.
• To provide a foundation for developing advanced optimization algorithms for high-precision QTL localization.

Main Methods:

• Explored two approaches for obtaining continuous marker-bracket IBD functions.
• Re-implemented an existing deterministic IBD estimation method to generate IBD functions.
• Developed a general IBD function approximation algorithm using existing IBD matrices.

Main Results:

• The re-implemented method achieved over a 2-fold improvement in computational efficiency and reduced memory usage for genome-wide IBD.
• The approximation algorithm generated IBD functions yielding accurate QTL variance component estimates.
• Storing IBD as polynomial functions significantly reduced memory requirements in QTL genome scans.

Conclusions:

• Continuous IBD function estimation offers substantial computational and memory efficiency gains for QTL mapping.
• These methods provide immediate utility for single QTL analyses and are particularly beneficial for complex genome scans involving multiple interacting QTL.
• The developed techniques are key to enabling widespread adoption of efficient optimization algorithms for precise QTL localization.