Genome-wide association analysis using multiple Atlantic salmon populations.

Area of Science:

• Aquaculture and Animal Genetics
• Statistical Genetics
• Genomic Prediction

Background:

• Previous studies showed low linkage disequilibrium (LD) phase persistence across Atlantic salmon breeding populations.
• Combining populations did not improve accuracy for genomic prediction in prior work.
• This study investigates combining populations for genome-wide association studies (GWAS) to enhance detection power and quantitative trait loci (QTL) mapping precision.

Purpose of the Study:

• To assess if combining Atlantic salmon populations improves detection power and mapping precision in GWAS.
• To compare mega-analysis (requiring individual data) with meta-analysis (using summary statistics).
• To evaluate methods for identifying independent or secondary signals, including conditional association analysis, COJO, and clumping.

Main Methods:

• Genome-wide association studies (GWAS) were performed within and across populations.
• Mega-analysis and various meta-analysis approaches were compared.
• Conditional association analysis, Approximate Conditional and Joint Analysis (COJO), and clumping were used to identify independent signals.

Main Results:

• Mega-analysis increased GWAS detection power (p-value reduction) and mapping precision compared to within-population analyses.
• COJO and clumping detected 1-19 QTL, while conditional analysis identified only one.
• Meta-analysis methods showed high correlation with mega-analysis results but yielded higher power with reduced precision.

Conclusions:

• Combining multiple datasets/populations via mega-analysis enhances GWAS detection power and mapping precision.
• Meta-analysis provides higher detection power than mega-analysis.
• Caution is advised when interpreting meta-analysis results due to potential inflation from population structure or cryptic relatedness.