This process yielded a model of how the OVOLs interact with the other TFs to influence OI MET. We tested this model for association with BC, PC, cancer, and MET in the lit erature and found it to be even more etc enriched than the OI MET model. This result is consistent with the hypoth esis that the OI MET TF model is also useful in under standing the impact of the OVOLs in MET and more generally in cancer, as well as how the OVOLs interact with the other four TFs in this process. By developing an improved understanding the genes, interactions, and related mechanisms impacting disease, we open up the possibility of intervening in disease progression. We used the OI MET TF model to understand how known drug/ gene interactions could impact the model and offer priori tized options for intervention.

We reflected our inference from the OI MET TF model back to the larger set of genes in the OI MET signature and tested this gene set for potential regulation by these TFs. In the OI MET gene set, we found signifi cant enrichment for binding motifs for the AP1/MYC pair. To investigate potential binding at these sites, we used publicly available ChIP Seq data to first test the hypothesis that AP1 binds preferentially in MET and Non MET cancer models, relative to a non cancer model. We also compared AP1 binding in the MET versus Non MET models. Results of these tests are consistent with AP1 acting in both the MET and Non MET can cer models. We then tested for preferential binding of the AP1/MYC pair, and again saw results consistent with this pair acting in both MET and Non MET cancer models.

While AP1 and MYC have long been associated with cancers, to our knowledge this is the first large scale test of the hypothesis that these TFs bind preferentially in cancer versus non cancer models for cancer related genes, and that they cooperate in binding. Taken together with evidence that FOS and JUN show differential expression in response to the OVOLs, these results are consistent with a regulatory cascade posed by the OI MET TF model. We also must consider the possibility that the OVOLs function in ways that are not specific to MET. This result has been seen with other transcription factors that were initially thought to act primarily in MET but were also found to impact cancer in ways not specific to MET. Conclusions In this work, we explore the etiology of OVOL Induced MET, focusing on commonalities between prostate cancer and breast cancer models, to test the hypothesis that the OVOLs regulate MET in multiple cancers. We generate a common OI MET gene expression signature, consistent with a common under lying genetic etiology for Entinostat MET in PC and BC, and show that the OI MET gene set is significantly enriched for can cer, therefore BC, PC, and MET associated genes.