The Drosophila heart provides a relatively simple system for the analysis of gene regulatory networks (GRNs), as it comprises just two cell types - contractile cardial cells (CCs) and non-muscle pericardial cells (PCs). Moreover, many transcription factors (TFs) that regulate Drosophila heart development have been identified and shown to play conserved roles in mammals. On p. 878, Alan Michelson and co-workers take a machine learning approach to identify new cardiac enhancers. The methodology, which integrates chromatin immunoprecipitation data with TF motif mapping, is highly successful at predicting enhancers that drive expression in cardiac cells, including PC- or CC-specific enhancers. Among the highly over-represented TF motifs in putative cardiac enhancers are motifs for the Myb and Su(H) TFs; the authors show that Myb regulates progenitor cell division, while Su(H) controls the PC/CC lineage decision. This work demonstrates the utility of machine learning to identify novel players regulating organ formation, and to piece together the GRNs underlying cell fate specification.