The nature of the methylation abnormalities undergone by cancer cell genomes is of great importance to cancer research, but current methods of methylation profiling are biased, inefficient, very expensive, and provide only partial coverage. The new method presented here combines new methods of fractionation of DNA according to methylation status, new methods of genomic DNA library construction, and ultra-high throughput DNA sequencing to allow efficient whole-genome methylation profiling even when only nanogram amounts of DNA are available. As the cost of whole-genome methylation profiling falls and the procedure is automated, it is easy to imagine that whole-genome methylation profiles could be used in the clinic for diagnostic and prognostic purposes in the foreseeable future. Another very important result of this efficient new method of whole-genome methylation profiling will be the first complete description of the actual genomic methylation pattern in normal and cancer cells. This will greatly improve our understanding of the changes that methylation patterns undergo in mammary carcinoma. It is known that human DNA is modified by the methylation of cytosine bases, and that the patterns of methylated cytosines are abnormal in cancer. The extent of the abnormalities is not understood because methylation patterns are difficult to discern. We introduce new technologies that will profile the methylation pattern of the entire genome as a means to identify methylation pattern abnormalities that are associated with human breast cancer. ? ? ?
