Analysis procedure. Sequence reads were aligned to the reference sequence (bosTau4) by the MAQ software. SNPs were called and filtered by MAQ and custom scripts, resulting in a final set of 2.44 million SNPs. Comparison with 25,726 array-based genotpyes revealed a false-negative detection rate of 49%. A false-positive detection rate of 1.1% was determined by comparison with 196 randomly selected SNPs genotyped with MALDI-TOF spectroscopy. By determining the false-positive detection rate in 75 coding SNPs with high coverage (≥16), we found evidence that the high false-positive detection rate in these SNPs is due to mapping errors caused by duplications that are not reflected in the reference sequence rather than to sequencing errors.
Eck et al. Genome Biology 2009 10:R82 doi:10.1186/gb-2009-10-8-r82