Rates of assay success and genotyping error when single nucleotide polymorphism genotyping in non-model organisms: a case study in the Antarctic fur seal

• Published in: Molecular Ecology Resources
• Main Authors: Hoffman, J.I., Tucker, R., Bridgett, S.J., Clark, M.S., Forcada, J., Slate, J.
• Format: Article in Journal/Newspaper
• Language: unknown
• Published: Blackwell 2012
• Subjects: Antarctic; The Antarctic; Antarc*; Antarctic Fur Seal; Arctocephalus gazella
• Online Access: http://nora.nerc.ac.uk/id/eprint/20166/; http://onlinelibrary.wiley.com/doi/10.1111/j.1755-0998.2012.03158.x/abstract

Although single nucleotide polymorphisms (SNPs) are increasingly being recognized as powerful molecular markers, their application to non-model organisms can bring significant challenges. Among these are imperfect conversion rates of assays designed from in silico resources and the enhanced potential for genotyping error relative to pre-validated, highly optimized human SNPs. To explore these issues, we used Illumina’s GoldenGate assay to genotype 480 Antarctic fur seal (Arctocephalus gazella) individuals at 144 putative SNPs derived from a 454 transcriptome assembly. One hundred and thirty-five polymorphic SNPs (93.8%) were automatically validated by the program GenomeStudio, and the initial genotyping error rate, estimated from nine replicate samples, was 0.004 per reaction. However, an almost tenfold further reduction in the error rate was achieved by excluding 31 loci (21.5%) that exhibited unclear clustering patterns, manually editing clusters to allow rescoring of ambiguous or incorrect genotypes, and excluding 18 samples (3.8%) with unreliable genotypes. After stringent quality filtering, we also found a counter-intuitive negative relationship between in silico minor allele frequency and the conversion rate, suggesting that some of our assays may have been designed from paralogous loci. Nevertheless, we obtained over 45 000 individual SNP genotypes with a final error rate of 0.0005, indicating that the GoldenGate assay is eminently capable of generating large, high-quality data sets for non-model organisms. This has positive implications for future studies of the evolutionary, behavioural and conservation genetics of natural populations.