Abstract In recent years, enormous effort has been dedicated to transcriptomic profiling of various physiological and pathological conditions, in particular in cancer biology field. Microarray gene expression (mRNA) datasets of thousands of human tumors are now publicly available and can be used to get insights of cancer processes by systems-based analysis and to identify biomarkers for improvement of cancer therapy. However, the high complexity of human transcriptome may be difficult to handle. In this regard, shifting the attention to the miRNome could be an advantage, since its complexity is at least 20-fold lower than that of a reference transcriptome (∼1000 miRNAs vs. ∼20,000 genes). Importantly, patterns of distinct miRNA expression profiles were observed in tumors and there is a growing interest in miRNAs, behaving as potential cancer determinants and biomarkers. Knowing that almost 50% of human miRNA genes are located within introns of coding genes and that they usually share transcriptional regulation, those publicly available datasets are likely to contain “latent” information on intronic-miRNA expression. In other words, it could be possible to predict the regulation of intronic-miRNA expression by simply analyzing their host genes profile (miR-HG). The aim of our work is to take advantage of cancer datasets, focusing mainly on breast cancer datasets, to provide proof of principle evidences that meta-analysis of miR-HG expression profiles can pinpoint intronic-miRNAs with a role in breast cancer cell biology, and a potential utility as cancer biomarkers. Using this approach, we successfully discovered a diagnostic microRNA signature enabling the identification of breast cancer molecular subtypes. Importantly, qRT-PCR analysis of just three intronic-miRNAs, using FFPE samples, was sufficient to identify more aggressive breast tumor subtypes (i.e. basal, HER2 and luminal B subtypes), with a ∼80% of accuracy, in patients with moderately differentiated breast cancer. Unexpectedly, in a number of cases, the regulation of expression of intronic-miRNAs was more relevant to cancer phenotypes than the expression of their host genes. In line with these encouraging results, we propose our data mining strategy as a valid tool for cancer research and other biomedical fields. Citation Format: Simona Monterisi, Giovanni D'Ario, Elisa Dama, Nicole Rotmensz, Stefano Confalonieri, Chiara Tordonato, Flavia Troglio, Giovanni Bertalot, Patrick Maisonneuve, Giuseppe Viale, Francesco Nicassio, Pier Paolo Di Fiore, Fabrizio Bianchi. Mining cancer gene expression databases for latent information on intronic microRNAs. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 233. doi:10.1158/1538-7445.AM2015-233