Methodology in phenome-wide association studies: a systematic review

Area of Science:

• Genetics and Genomics
• Biostatistics
• Computational Biology

Background:

• Phenome-wide association studies (PheWAS) are crucial for discovering novel genetic associations across diverse phenotypes.
• Understanding the methodology, advantages, and challenges of PheWAS is essential for advancing genetic research.

Purpose of the Study:

• To systematically review and summarize the methodology of PheWAS.
• To discuss the advantages and challenges associated with PheWAS.
• To provide insights into the implications for future PheWAS studies.

Main Methods:

• Systematic literature search of MEDLINE and EMBASE databases up to April 24, 2021.
• Inclusion of 195 eligible articles from an initial 1035 identified studies.
• Extraction and summarization of PheWAS methodology, including analysis techniques and software tools.

Main Results:

• The majority of included studies (77.0%) had over 10,000 participants.
• Electronic medical records (92.1%) were commonly used for phenome definition, with genetic variants (78.7%) as predictors.
• Replication analysis was performed in 41.0% of studies, yielding consistent results in 94.5% of cases.

Conclusions:

• A comprehensive workflow for PheWAS was established, detailing critical steps from quality control to result visualization.
• This review enhances understanding of PheWAS design, aiding in the identification of understudied or overstudied outcomes.
• Guidance is provided on selecting appropriate software and tools for diverse PheWAS data structures.