Virtually all methods for the detection of mutations (polymorphism or variant) rely on polymerase chain reaction (PCR). Direct
sequence determination of a PCR product is the gold standard for identifying mutations. However, the vast majority of the
signal in the sequencing data is derived from nonvariant sequence, and can be a source of noise. Thus, a somewhat high false
positive rate is inevitable when rare mutations are searched for in a large genomic region or in a region of many genomes.
Techniques to detect variants as positive signals have the advantage of intrinsically low false positive rate, and are suitable
methods to preselect fragments that carry mutations among an excess of nonmutated fragments. Such techniques are especially
useful, for example in surveying for possible mutations in genes suspected to be responsible for genetic diseases, or finding
single-nucleotide polymorphisms (SNPs) in blindly amplified genomic segments.