Human genomic data have different statistical properties than the data of randomised controlled trials

Area of Science:

• Genetics
• Bioinformatics
• Statistical Genetics

Background:

• The analogy between Mendelian gene reshuffling and randomized controlled trials (RCTs) has been proposed.
• This perspective suggests genetic variation arises from a process similar to experimental randomization.

Purpose of the Study:

• To critically evaluate the argument that Mendelian reshuffling of genes and genomes is analogous to randomized controlled trials.
• To identify and discuss the limitations of this analogy in the context of genetic research.

Main Methods:

• Critical analysis of the proposed analogy between genetic processes and experimental design.
• Examination of assumptions underlying randomized experiments in light of genetic principles.
• Review of potential challenges and limitations in applying RCT concepts to genetic variation, including genome-wide association studies (GWAS).

Main Results:

• The proposed analogy faces significant inconsistencies regarding the demands of randomized experiments.
• Disequilibrium across chromosomes contradicts the assumption of statistical independence required for RCTs.
• Genome-wide association studies (GWAS), a key method in genetic research, exhibit notable pitfalls, such as low repeatability.

Conclusions:

• The analogy between Mendelian reshuffling and randomized controlled trials is not convincing due to inherent inconsistencies and conflicting assumptions.
• The practical application of RCT principles to genetic variation is limited by factors like linkage disequilibrium and methodological challenges in genetic studies.