Test-statistic inflation in methylome-wide association studies

Area of Science:

• Genomics
• Epigenetics
• Bioinformatics

Background:

• Methylome-wide association studies (MWAS) are increasingly common for identifying genetic associations with DNA methylation.
• Existing quality control (QC) pipelines may not sufficiently remove technical artifacts, leading to inflated test statistics in MWAS.

Purpose of the Study:

• To investigate the impact of commonly used QC pipelines on test statistic inflation in MWAS.
• To demonstrate the effectiveness of an improved QC pipeline in reducing false positives.
• To advocate for standardized reporting of the test-statistic inflation parameter lambda in MWAS.

Main Methods:

• Reanalysis of Gene Expression Omnibus (GEO) datasets using an enhanced QC pipeline.
• Comparison of test-statistic inflation parameter lambda before and after improved QC.
• Quantification of the reduction in methylome-wide significant findings post-QC.

Main Results:

• The improved QC pipeline reduced the mean/median lambda from 20.16/15.17 to 3.07/1.14.
• A significant reduction in methylome-wide significant loci was observed, with a mean/median decrease of 65,688/57,805.
• These findings indicate substantial removal of false positives through more thorough QC.

Conclusions:

• Inadequate QC in MWAS can lead to significant test statistic inflation and false discoveries.
• More extensive QC is crucial for accurate results in MWAS.
• Standardizing the reporting of the lambda inflation parameter will enhance the reliability and interpretability of MWAS findings.