Cancer cells have abnormal gene expression patterns, however, the transcription factors and the architecture of the regulatory network that drive cancer specific gene expression profiles is often not known. Here we studied a model of Ras-driven invasive tumorigenesis in Drosophila larval epithelial tissues and combined in vivo genetic analyses with high-throughput sequencing and computational modeling to decipher the regulatory logic of tumor cells. Surprisingly, we discovered that tumor specific gene expression is driven by a highly interconnected network composed of few transcription factors. These are: Stat, Mef2, the AP-4 homolog Cropped, the nuclear receptor Ftz-f1, the bHLH factors Myc and Taiman, and the AP-1 transcription factors Kayak, ATF-3, Pdp1, and dCEBPG. Many of these transcription factors are ectopically expressed and/or hyperactivated in human tumors. The members of this tumor master regulatory network are predicted to directly regulate the majority of the tumor specific gene expression profile. Similar to networks of master regulators that control organ development and cellular differentiation, there is a predicted high degree of co-regulation of target genes, and these network members are required in multiple eptihelia for tumor growth and invasiveness. We further found that Yki/Sd and bZIP/AP-1 factors, the downstream transcription factors of the Hippo and JNK pathways, initiate cellular reprogramming by activating several transcription factors of this network. Thus, modeling regulatory networks identified an ectopic yet highly ordered network of master regulators that control cancer cell specific gene expression. RNA-seq gene expression profiling across Drosophila 3rd instar larval imaginal discs (eye-antenna, wing and leg) in a hh driven tumor model, perturbations and controls.