Bioinformatics for miRNA Analysis
One of the most powerful tools of all for analyzing mRNA is the computer. Using specialized algorithms, researchers can analyze gene sequences for miRNAs and predict miRNA targets and pathways. Since unbiased miRNA analysis is difficult to impossible in vitro, a powerful software package to narrow down the hypotheticals is indispensible.
Actigenics (www.actigenics.com), now a division of Cepheid (www.cepheid.com), offers a comprehensive miRNA analysis platform (MiRgate).
Actigenics’ miRNA analysis begins with bioinformatic predicitions of thousands of novel miRNA candidates, followed by a custom microarray analysis to create a microarray expression profile for clinical samples. Then the MiRgate software suite correlates microarray results with significant pathways (using the company’s extensive database), producing a custom report for each sample. MiRgate will also analyze noncoding DNA for miRNAs, search for SNPs, and predict miRNA target sequences.
There is considerable affinity between Cepheid and Actigenics, explains David Persing, M.D., Ph.D., chief medical and technology officer and executive vice president for Cepheid. “We managed to acquire the expertise of a group that was extremely talented at being able to predict the presence of miRNA candidates within the human genome.
“Dr. Bernard Michot, formerly of INSERM, joined us as head of the group in the Cepheid France facility. He and his group of bioinformatics specialists developed a software package that is capable of producing accurate predictions of miRNA candidates and potential diagnostic markers not yet represented within the public domain data set.”
For Cepheid, this was an opportunity to access proprietary diagnostic targets for a broad array of diseases, and for Actigenics, it was an opportunity to commercialize a unique technology.
“We have reasons to believe miRNA markers might be among the most accurate for the diagnosis of cancer, inflammation, and autoimmune conditions such as systemic lupus, and there are a lot of areas yet to be explored by our system,” notes Dr. Persing. “What we expect to get out of it in practical terms is a short list of diagnostic markers that we can incorporate into tests on our GeneXpert diagnostic platform.”
In some cases, Cepheid has identified miRNA targets within regions of amplified and deleted chromosomes in cancers. For example, the region 17q12-22 is commonly amplified in breast cancer. It contains the gene for ErbB2, which is the target of Herceptin.
Cepheid predicted that two miRNAs from the same region would be overexpressed in breast cancer with potential regulatory roles apart from ErbB2 amplification. Overexpression of these two molecules was later demonstrated on microarrays. Interestingly, no miRNAs from the public domain data set were found to map to this region of the chromosome. These miRNAs may themselves be good diagnostic markers and they may in turn point to downstream pathways that may also be good candidates for new companion diagnostics in the mold of Herceptin/ErbB2.
Happily, the greatest obstacle to full exploitation of miRNA biology has been removed—that of simply not knowing it existed. miRNAs have begun to explain mysteries such as the mechanisms of embryonic development and the haywire genetics of cancer. And yet, the mind-boggling complexity of tiny regulatory elements that control hundreds of genes at a time may make us nostalgic for the days when noncoding DNA was junk and gene regulation consisted of a comfortably large protein stationed like a parking lot attendant on top of a very obvious promoter region.
But as with most nostalgia, we’d have to admit that the world is a more interesting place with miRNA in it, and as better tools become available for analyzing miRNA, those dizzying networks of interactions should slowly begin to make sense.