One of the current challenges of oncogenomics is to elucidate which of the aberrations observed in the cancer genome is involved in the disease and which of them are just bystanders occurring stochastically due to the cell genomic instability, i.e. to distinguish driver events from passengers. This is indeed demanding when analyzing copy number alterations (CNAs), in which very large regions of the DNA (and thus a large number of genes) may be affected. The current approach for dealing with this issue is to assess the recurrence of the alteration across multiple samples, since those alterations occurring more than expected by chance should point out significant events in terms of the disease. However, this exhibits some drawbacks, as the underestimation of low-recurrent drivers, the difficulty of assessing the background model and, finally, that it does not take into account how the alteration may impact the normal behavior of the gene.
Therefore, we have developed Oncodrive-CIS, a novel method to measure the in cis effect of copy number changes aimed to identify those involved in the tumorigenesis. The main rationale of our method is that those events driving the disease through CNAs are more prone to be biased towards a larger in cis change than passengers. You can find the details of the method implementation in the manuscript publication; in few words, what Oncodrive-CIS calculations retrieve is a list of genes ranked according to their in cis expression change due to CNAs.
The performance of the method has been evaluated by two ways: first, by using a synthetic data generator developed by Louhimo R et al., Oncodrive-CIS demonstrated better accuracy in detecting putative association between gene dosage and expression than any other existing method aimed for this purpose. Second, we have analysed two real cancer data sets retrieved from the Cancer Genome Atlas Data Portal (Gliobastoma multiforme and Ovarian serous carcinoma), and Oncodrive-CIS resulted successful in selecting genes known to drive cancer or likely candidates of leading to tumor phenotypes.
Oncodrive-CIS follows the principles of Oncodrive-FM, another method developed in our lab, in the sense that both of them assess the role of somatic aberrations according to the functional impact they cause in the gene: the former does it by assessing the in cis expression change caused by CNAs, the latter by evaluating the impairment caused in the protein function by somatic mutations. Both of them also share the feature of being scalable to the input data, a matter of concern taking into account the amount of ongoing data coming from large cancer genomic studies.
In summary, we conclude that Oncodrive-CIS is a method that should be taken into account when elucidating the role of CNAs in driving cancer, although it may be considered not exclusive from methods based on other criteria as the aforementioned recurrence analysis. Oncodrive-CIS integrates data from gene dosage and expression, and at present we are working for including methylation data as another source of in cis changes.
Tamborero D, Lopez-Bigas N, Gonzalez-Perez A (2013) Oncodrive-CIS: A Method to Reveal Likely Driver Genes Based on the Impact of Their Copy Number Changes on Expression. PLoS ONE 8(2): e55489. doi:10.1371/journal.pone.0055489 [ Download PDF ]
How to run:
Oncodrive-CIS web contains the code and documentation to run it (http://bg.upf.edu/oncodrive-cis)
How to identify cancer drivers from tumor somatic mutations