Weeding out false positives in the hunt for cancer genes

21 June 2013

New research published in Nature has shed light on the so-called ‘fishy genes’ in cancer – genetic mutations consistently linked with cancer, but which appear to have wholly unrelated functions.

Cancer genomics is an important field; identifying the key ‘driver’ mutations involved in the initiation, progression and spread of tumours gives insight into their true biological nature, informing the development of new treatments. However, studies to identify genes linked with cancer also typically pinpoint some with no plausible link to the disease – for example, olfactory receptor genes and lung cancer. Researchers began to postulate that there was, in fact, no direct link, but certain genes recurred in results simply because they are unusually prone to mutation.
 
Mutations in these genes may represent a sort of pattern associated with a given type or types of cancer – for example, genes susceptible to mutations induced by tobacco are commonly linked with lung cancer – but they do not play a role in formation of the tumours. A new analytical methodology to identify genes truly linked with cancer (MutSigCV) specifically excludes genes in these highly mutable genomic regions, which tend to be positively selected for in cancer cells, drastically reducing the number of true candidate genes.
 
Scientists from The Cancer Genome Atlas (TCGA) project who developed the tool say that it will now be possible to analyse large cancer datasets, excluding the highly mutable genes, in order to pinpoint those with a genuine role in the disease.

More from us