MicroRNAs (miRNAs) and genes work cooperatively to form the kernel part of gene regulatory system and affect many crucial biological processes. However, the detailed combinatorial roles of most miRNAs and genes in cellular processes and diseases are still unclear. The huge amount of diverse functional genomic data provides unprecedented opportunities to study the miRNA-gene co-regulations. How to integrate diverse genomic data to identify the regulatory modules of miRNAs and genes is a challenging problem in computational biology. Recently, we have proposed a mathematical data integration framework to discover the miRNA-gene regulatory co-modules. We have applied the proposed method to integrate a set of heterogeneous data sources including the expression profiles of miRNAs and genes on 385 human ovarian cancer samples as well as miRNA-gene interactions and gene-gene interactions. The revealed co-modules show significant biological relevance and potential associations with ovarian cancers and others.