Methods and compositions are provided for detecting the presence of nucleic acid sequence variants in a subpopulation of nucleic acid molecules in a biological sample. These methods are particularly useful for identifying individuals with mutations indicative of cancer.
