PMERGE: Computational filtering of paralogous sequences from RAD-seq data.

Area of Science:

• Population Genetics
• Genomics
• Bioinformatics

Background:

• Restriction-site associated DNA sequencing (RAD-seq) is valuable for population genetics.
• Distinguishing paralogous sequence variants (PSVs) from true single-nucleotide polymorphisms (SNPs) is a challenge in de novo RAD-seq without a reference genome.
• The impact of PSVs on downstream population genetics analyses is not well understood.

Purpose of the Study:

• To introduce PMERGE, a network-based approach for identifying PSVs in de novo RAD-seq data.
• To assess the effectiveness of PMERGE in distinguishing PSVs from orthologous SNPs.
• To evaluate the impact of PSV filtering on population structure inference.

Main Methods:

• Developed a network-based algorithm (PMERGE) to connect DNA fragments based on sequence similarity.
• Applied PMERGE to de novo RAD-seq data from Atlantic salmon (Salmo salar) and green crab (Carcinus maenas).
• Compared PMERGE-identified PSVs with those found through genome alignment and assessed effects on population structure.

Main Results:

• PMERGE identified 87% of PSVs in Atlantic salmon and 62% in green crab, comparable to genome alignment.
• Removal of identified paralogs significantly altered inferred population structure in Atlantic salmon.
• PMERGE is compatible with the Stacks analysis package.

Conclusions:

• PMERGE effectively identifies PSVs in de novo RAD-seq data, crucial for accurate population genetics.
• Filtering PSVs using PMERGE can significantly impact population structure inferences.
• The method provides a valuable tool for genetic marker analysis in non-model organisms.