At the co-regulatory network level, functional modules comprising microRNAs (miRNAs), transcription factors (TFs), and common target genes (mRNAs) are found to be widespread in diverse organisms from bacteria to human, suggesting that co-regulatory functional modules (CRMs) serve as basic building blocks of transcription networks. Identification of CRMs would contribute to explore the miRNAs and TFs regulatory mechanism to specific cancer. However, few studies considered the functional modules which contain TFs and overlapping parts now. In this paper, we propose a novel computational framework called overlapping spectral clustering (OSC) to systematically detect overlapping CRMs using miRNA/mRNA/TF expression profiles. First, empirical Bayes theories are used to improve building more exact relation matrix instead of simple Pearson correlation analysis. In order to ensure the adaptivity of the whole framework, the eigenvalue decomposition (Eigengap) is employed to determine the module number automatically. Then considering key regulators which may involve in more function modules, we propose a novel overlapping detection approach to observe whether the edges between modules can be overlapped by other modules. Comparing with existing methods on breast cancer and ovary cancer data sets from The Cancer Genome Atlas, we showed that the CRMs identified by OSC are more functionally enriched. 80% TFs and 90% miRNAs in CRMs are related to corresponding diseases. And the overlapping parts between CRMs were also analyzed and verified by standard databases and previous literatures.