Sex-Specific Polygenic Risk Scores and Replication in a Model-Free Analysis of Schizophrenia Data

Area of Science:

• Genetics
• Psychiatric Disorders
• Biostatistics

Background:

• Common heritable traits like schizophrenia are influenced by numerous genetic variants, each with a small effect.
• Polygenic risk scores (PRS) are favored for combining effects of multiple variants for genetic prediction.
• Existing PRS methods primarily aim for phenotype prediction.

Purpose of the Study:

• To develop and validate allelic polygenic risk scores for improved genetic prediction of schizophrenia.
• To account for potential sex-specific genetic heterogeneity in schizophrenia risk.
• To assess the accuracy and replicability of PRS in independent male and female datasets.

Main Methods:

• Separated male and female datasets to develop independent allelic polygenic risk scores.
• Focused on highly predictive genetic variants rather than those with high statistical significance.
• Developed a statistical test for cross-sex replication and significance assessment.

Main Results:

• High accuracy (>95%) in phenotype prediction was achieved using as few as 5000 highly predictive variants.
• Misclassification rates were low: 2.8% in females and 3.3% in males.
• 544 disease-associated variants showed significant shared associations between sexes, indicating robust replication.

Conclusions:

• Allelic polygenic risk scores based on individual genotypes provide highly accurate, cross-validated phenotype predictions.
• The method demonstrates effective genetic prediction of schizophrenia, accounting for sex differences.
• Identified significant shared genetic variants between males and females, supporting the robustness of the findings.