Originally published as Genetics Published Articles Ahead of Print on April 3, 2007.

Genetics, Vol. 176, 789-800, June 2007, Copyright © 2007
doi:10.1534/genetics.106.067843

SFP Genotyping From Affymetrix Arrays Is Robust But Largely Detects Cis-acting Expression Regulators

Z. W. Luo*,{dagger}, E. Potokina*, A. Druka{ddagger}, R. Wise§, R. Waugh{ddagger} and M. J. Kearsey*,1

* School of Biosciences, University of Birmingham, Birmingham B15 2TT, United Kingdom, {ddagger} Scottish Crop Research Institute, Invergowrie, Dundee DD2 5DA, Scotland, United Kingdom, {dagger} Laboratory of Population and Quantitative Genetics, Department of Biostatistics, School of Life Sciences, Institute of Biomedical Sciences, Fudan University, Shanghai 200433, China and § Corn Insects and Crop Genetics Research, USDA–ARS, Department of Plant Pathology and Center for Plant Responses to Environmental Stresses, Iowa State University, Ames, Iowa 50011-1020

1 Corresponding author: School of Biosciences, University of Birmingham, Birmingham B15 2TT, United Kingdom.
E-mail: m.j.kearsey{at}bham.ac.uk

The recent development of Affymetrix chips designed from assembled EST sequences has spawned considerable interest in identifying single-feature polymorphisms (SFPs) from transcriptome data. SFPs are valuable genetic markers that potentially offer a physical link to the structural genes themselves. However, most current SFP prediction methodologies were developed for sequenced species although SFPs are particularly valuable for species with complex and unsequenced genomes. To establish the sensitivity and specificity of prediction, we explored four methods for identifying SFPs from experiments involving two tissues in two commercial barleys and their doubled-haploid progeny. The methods were compared in terms of numbers of SFPs predicted and their ability to identify known sequence polymorphisms in the features, to confirm existing SNP genotypes and to match existing maps and individual haplotypes. We identified >4000 separate SFPs that accurately predicted the SNP genotype of >98% of the doubled-haploid (DH) lines. They were highly enriched for features containing sequence polymorphisms but all methods uniformly identified a majority of SFPs (~64%) in features for which there was no sequence polymorphism while 5% mapped to different locations, indicating that "SFPs" mainly represent polymorphism in cis-acting regulators. All methods are efficient and robust at predicting markers for gene mapping.