ABSTRACT
Despite much discussion of the importance of quantifying and reporting genotyping error in molecular studies, it is still not standard practice in the literature. This is particularly a concern for amplified fragment length polymorphism (AFLP) studies, where differences in laboratory, peak-calling and locus-selection protocols can generate data sets varying widely in genotyping error rate, the number of loci used and potentially estimates of genetic diversity or differentiation. In our experience, papers rarely provide adequate information on AFLP reproducibility, making meaningful comparisons among studies difficult. To quantify the extent of this problem, we reviewed the current molecular ecology literature (470 recent AFLP articles) to determine the proportion of studies that report an error rate and follow established guidelines for assessing error. Fifty-four per cent of recent articles do not report any assessment of data set reproducibility. Of those studies that do claim to have assessed reproducibility, the majority (~90%) either do not report a specific error rate or do not provide sufficient details to allow the reader to judge whether error was assessed correctly. Even of the papers that do report an error rate and provide details, many (≥23%) do not follow recommended standards for quantifying error. These issues also exist for other marker types such as microsatellites, and next-generation sequencing techniques, particularly those which use restriction enzymes for fragment generation. Therefore, we urge all researchers conducting genotyping studies to estimate and more transparently report genotyping error using existing guidelines and encourage journals to enforce stricter standards for the publication of genotyping studies.
