In the search for molecular markers which can serve for early detection of cancer, for gauging prognosis, or assessing cancer risk, abnormal methylation of gene promoter regions is one of the most promising strategies at present. Unlike mutations, which can often occur at multiple sites in a gene, thus requiring multiple primers for PCR strategies to detect a change in any given patient, these methylation changes always occur at the same starting points in the gene. Thus they can be targeted with a single PCR primer set that would be suitable for any patient. This approach, using DNA from sites such as sputum, serum, etc. looks very promising in multiple proof of principle studies. This being said, one of the next steps would be to provide full coverage of the genomes of multiple cancer types for hypermethylated genes. The paper shows how that will probably be possible by demonstrating that one of just four markers for each tumor type studied—breast, colon, kidney, lung, leukemias, etc.—will be positive in 60 to 80% for each cancer. Thus panels can be constructed to serve for each major cancer. This is what I believe is serving to provide for interest in the paper. At a time where there is a great enthusiasm for using microarrays and other means for molecularly "profiling" tumors for patterns of markers etc., our approach provides a relatively simple but potentially powerful approach which could be clinically practical. We have another article, which has recently appeared online in the journal Nature Genetics, which describes a way to screen for these hypermethylation markers, and which shows derivation of a panel that covers 99% of primary colon cancers studied.