To determine promoter and genome-wide patterns of methylation alterations in AML and high-risk sarcoma samples from known clinical/molecular subgroups. We hypothesize that tumor samples with known adverse prognostic features will have distinct regional changes in CpG methylation when compared to lower-risk counterpart samples. To test this hypothesis, we are developing novel methods to prepare detailed, genome-wide maps of methylation in high risk pediatric cancer samples using a novel method of enrichment of methylated followed by high-resolution genomic tiling microarray analysis.