Mendelianization: Concentrating Polygenic Signal Into a Single Causal Locus

Area of Science:

• Genetics
• Biostatistics
• Computational Biology

Background:

• Complex disorders involve numerous genetic variants, hindering mechanistic understanding.
• Classical Mendelian diseases, caused by single loci, offer clearer interpretability.
• Current methods struggle with the polygenicity of common diseases.

Purpose of the Study:

• Introduce Mendelianization, an algorithm to identify single causal loci for complex traits.
• Enhance the interpretability of genetic contributions to pathophysiology.
• Improve statistical power for detecting genetic influences on disease.

Main Methods:

• Developed a novel algorithm named Mendelianization.
• Learns weighted combinations of outcomes to aggregate phenotypes.
• Leverages summary statistics (z-scores) for analysis, handling partial sample overlap.

Main Results:

• Mendelianization concentrates genetic associations at a single locus.
• Proven causal under four natural genetic data assumptions.
• Demonstrated enhanced statistical power in simulations for complex disorders.

Conclusions:

• Mendelianization offers a powerful approach to dissecting genetic architectures of complex diseases.
• The method facilitates clearer interpretation of genetic variants' roles in pathophysiology.
• Applicable to heterogeneous disorders like major depression and alcohol use disorder.