Leveraging haplotype information in heritability estimation and polygenic prediction

Area of Science:

• Genetics
• Computational Biology
• Psychiatry

Background:

• Polygenic risk scores (PRS) are crucial for precision medicine and psychiatry but face challenges with diverse populations due to reliance on external summary statistics.
• Current PRS methods assume matching data distributions, potentially limiting their accuracy in populations with differing genetic architectures.

Purpose of the Study:

• To introduce hapla, a novel framework for reference-free polygenic prediction using haplotype clustering.
• To improve the accuracy and robustness of heritability estimation and PRS in complex traits.

Main Methods:

• Developed a novel algorithm for clustering haplotypes in phased genotype data.
• Utilized inferred haplotype clusters for reference-free polygenic score estimation.
• Applied the method to a simulation study and the iPSYCH2012 case-cohort for depression and schizophrenia.

Main Results:

• The hapla framework accurately estimates heritability and performs polygenic prediction.
• Haplotype-based polygenic scores demonstrate robust outperformance compared to standard genotype-based approaches.
• The method shows promise for application in diverse populations.

Conclusions:

• Haplotype-based reference-free prediction offers a more robust alternative to current polygenic risk score methods.
• This approach can overcome limitations associated with external summary statistics and population differences.
• The hapla framework has the potential to enhance clinical decision-making in precision medicine and psychiatry.