NEW YORK (GenomeWeb News) – New research is demonstrating the power of a computational algorithm for filtering genomic data so that scientists can detect the mutations most likely to have a role in specific cancers.
Researchers from the Dana-Farber Cancer Institute, Memorial Sloan-Kettering Cancer Center, and elsewhere developed the so-called genome topography scan, or GTS, algorithm to rank mutations associated with a type of brain cancer called glioblastoma and finger those mutations most likely to affect cancer development and progression. In the process, they identified a new functional pathway involved in glioblastoma’s tumorigenesis. The results appear online today in the journal Cancer Cell.
“We have demonstrated here that GTS can address one critical need in the development of a functional map of [glioblastoma] genetic targets: namely, to prioritize those genomic alterations that are likely to be of importance from among those that are more likely to be bystanders of the cancer process,” co-lead authors Ruprecht Wiedemeyer, a post-doctoral fellow at Dana-Farber, and Cameron Brennan, a neurosurgeon and researcher affiliated with Memorial Sloan-Kettering and Weill-Cornell Medical College, and colleagues wrote.
Although many genes have been implicated in different types of cancer, determining which mutations are really relevant to the condition being studied is often daunting.
The GTS approach involves combining several types of data about chromosomal copy aberrations found in cancerous cells — including the level of variation, how often they occurred, and whether they were specifically focused or covered a wide region on the chromosome.
For this study, the team applied this three-dimensional, high-resolution genome topography scan to glioblastoma, a type of adult brain cancer that is notoriously aggressive and difficult to treat.
First, they identified genes of interest in 28 related tumor samples and 18 cell lines using array-comparative genomic hybridization profiling on Agilent’s 60-mer 44K or 244K density microarrays.
They then analyzed the data to determine factors such as amplitude and width of copy number aberrations and their frequency in different parts of the genome. By putting this information into the algorithm, the team was able to rank the top 50 amplification and deletion mutations associated with glioblastoma.
Although they picked up some genes that had previously been associated with glioblastoma, the approach allowed them to uncover a new glioblastoma-related gene, called p18INK4C, whose depletion is linked to the cancer. As it turns out, this gene seems to function in a cellular pathway that’s parallel to another gene often missing from brain cancer cells: p16INK4A, a gene that codes for a tumor suppressor protein.
Subsequent experiments supported the notion that these two related genes function in parallel pathways, curbing glioblastoma. On the other hand, when both genes are absent or mutated, glioblastoma develops or becomes more dangerous. For instance, when the researchers used short hairpin RNAs to knock down p18INK4C in cells lacking p16INK4A, the cells became more tumorigenic.
“Just a few years ago, the view was that pathways were largely linear,” senior author Lynda Chin, a dermatologist and oncologist affiliated with the Dana-Farber Cancer Institute and the Belfer Institute for Innovative Cancer Science, said in a statement. “We’re increasingly coming to appreciate, however, that they operate in concert — that each one has multiple tentacles reaching out to other pathways and they function collectively as a network.”
And, researchers say, the algorithm may also prove useful for interpreting data collected from large-scale cancer genome research projects. It is also expected to be added to the BioConductor website, which hosts open-source bioinformatics software.