SNP discovery in nonmodel organisms: strand bias and base-substitution errors reduce conversion rates

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Single nucleotide polymorphisms (SNPs) are crucial genetic markers in various organisms.
• Current SNP discovery in nonmodel organisms often relies on filtering random sequencing errors.

Purpose of the Study:

• To assess the impact of systematic sequencing errors on SNP discovery in orange roughy (Hoplostethus atlanticus).
• To improve SNP discovery efficiency in nonmodel organisms.

Main Methods:

• Polymorphism identification using SAMtools on assembled genomic DNA from seven individuals.
• Filtering to minimize sequencing and assembly errors.
• SNP genotyping using an Illumina Infinium chip on 1734 individuals.
• Analysis of predictors for assayable SNP success (coverage, variant support, type, strand-bias, probe design score).
• Utilizing BLASTX for identifying single-copy genomic regions without a reference genome.

Main Results:

• Filtering systematic sequencing errors substantially increases SNP discovery efficiency.
• Five key predictors influence the success of obtaining assayable SNPs.
• BLASTX is effective for identifying single-copy regions in de novo assemblies.

Conclusions:

• Accounting for systematic errors is vital for efficient SNP discovery in nonmodel organisms.
• The findings provide a framework for developing robust SNP genotyping assays for species lacking reference genomes.