Generating evidence-backed hypotheses in silico for novel prostate cancer biomarkers, Part V

The story of how, within just a few weeks, a scientist can uncover plausible biomarkers for a complex disease using IPA® for in silico research. We invite you to hear Dr. Billaud present this in silico study via this pre-recorded video of his presentation.

By Jean-Noel Billaud, Associate Staff Scientist, Ingenuity. September 27, 2011

(View Part I, Part II, Part III, or Part IV)

Part V: Identifying the hypothetical biomarker miRNA-mRNA pairs
Final steps in selecting plausible prostate cancer biomarkers

Since this is the final installment of the series, I’d like to take this opportunity to summarize the in silico work done so far to identify possible prostate cancer biomarker. First, I leveraged data from two existing studies in IPA. The first study (Study 1) generated a gene signature comprised of a list of 158 genes. The authors describe this gene signature: “This AR-SRF signature is sufficient to distinguish microdissected benign and malignant prostate samples, and it correlates with the presence of aggressive disease and poor outcome.”  Study 2 demonstrated that, (citing the authors), “differential miRNAs in prostate cancer are useful diagnostic and prognostic indicators.”

I wanted to examine the overlap of Study 1 and Study 2. The microRNA Target Filter in IPA allowed me to associate mRNA of Study 1 with miRNA of Study 2 based on their expression level. I could then easily connect these two biological components into a meaningful network, in which miRNA-mRNA are linked by scientific findings in the Ingenuity® Knowledge Base. This network (See Figure 7) displayed in IPA allowed me to visualize the relationships and connections between these miRNA-mRNA pairs. The network of miRNA-mRNA pairs provided a basis for proposing new hypothetic biomarkers for PCa because they represent a unit of regulation of a particular gene. Thus, both factors (miRNA and mRNA) were taken into account to propose relevant biomarkers for this disease. I was able to select inverse pairing as an option in the microRNA Target Filter. These miRNAs and the targeted mRNAs have their expression inversely correlated.

For next steps, I drilled down into the underlying scientific findings in these networks (by clicking on the edges in IPA to review source material), to help me then to select and propose potential biomarkers (miRNA-mRNAs pairs). For instance “expression of mir-96 was associated with cancer recurrence after radical prostatectomy” and “prognostic information was confirmed by an independent tumor sample set from 79 patients” the corresponding two mRNA targets of mir-96, SOX6 and DIAPH2 are down-regulated. The pairs mir-96-SOX6 or mir-96-DIAPH2 could then represent a potential new biomarker.

The complete network was thus decorated with the genes and associated miRNAs that are plausible biomarkers. This was the basis for proposing 3 sets of microRNA-mRNA pairs as biomarkers: 1) one set of disease progression biomarkers for the metastatic status (miRNA181-F5; miRNA181 -GNA12; miRNA181-CRIM1); 2) one set of diagnostic biomarkers for primary tumors (miRNA181-CYR61; miRNA125-RAB3IP; miRNA145-CITED2); and 3) one set of prognostic biomarkers for onset of recurrence (miRNA96-SOX6; miRNA96-DIAPH2; miRNA181-AHNAK).

In conclusion, IPA helped me to easily visualize the Heemers et. al. mRNA expression data and Jung et. al. microRNA expression data and identify patterns of coordinated gene expression between microRNA-mRNA pairs. In addition, the Ingenuity Knowledge Base provided a wealth of contextual facts derived from primary publications and leading third party databases, including those on clinically validated biomarkers, which helped me to further investigate the potential for these microRNA-mRNA pairs as biomarkers. This study illustrates the value and impact of analyzing data using IPA for research in silico to identify potential biomarkers such as the microRNA-mRNA pairs identified here.

I invite you to attend my presentation live in the Ingenuity webinar series or listen to the recorded versions of the webinars to learn more about the scientific reasoning behind the selection of these plausible prostate cancer biomarkers.

Click here to go to the home page.