Posted on 05/12/2009 7:34:57 PM PDT by neverdem
Researchers question focus on coding regions.
COLD SPRING HARBOR
To help battle their way through the stream of data coming in from human gene sequencing, major cancer-genome screening projects such as the International Cancer Genome Consortium (ICGC) seem to be choosing to simplify matters.
The ICGC aims eventually to sequence the full genomes of 25,000 tumour samples as well as those of the people from whom the tumours were taken, which would give 50,000 distinct genomes.
But in the near term, the project is doing targeted sequencing of just the 1% of the genome known to code for proteins — the 'exons' within genes.
Sequencing of the 'exome' — all the exons in the genome — involves chopping the genome into millions of pieces and capturing and sequencing only selected DNA from exon regions. It differs from transcriptome sequencing by focusing on DNA rather than the expressed RNA in a given cell, and it promises to be vastly cheaper than whole-genome sequencing. It will be a significant focus of the ICGC, which comprises ten projects from nine member countries, says Tom Hudson of the Ontario Institute for Cancer Research in Toronto and a member of the ICGC secretariat.
Last week, at the 'Biology of Genomes' meeting at Cold Spring Harbor Laboratory in New York state, some cancer researchers questioned whether exome sequencing is the most efficient way forward. They say it could represent a piecemeal half-step, and not provide a full picture of the mutations that lead to cancer.
At the conference, Michael Stratton of the Wellcome Trust Sanger Institute in Cambridge, UK, presented early results from a study of 24 breast-cancer samples that analysed more than 2,000 chromosomal rearrangements, including regions in which vast tracts of DNA were duplicated, swapped between chromosomes, inverted or otherwise adulterated.
With so many potentially deleterious rearrangements occurring in any given cancer cell, it becomes difficult to distinguish what Stratton calls the "driver" mutations, which spur and maintain cancer development, from "passenger" mutations that are just along for the ride.
Drivers might be in the coding regions of the genome, but some will presumably be in regulatory elements and other non-coding sequences — meaning that whole-genome sequences will ultimately be necessary, he says.
Elaine Mardis, of Washington University in St Louis, Missouri, offered a glimpse of what else could get missed by focusing on the exome with current technologies. Building on her recent whole-genome sequences of a patient with acute myeloid leukaemia (T. J. Ley et al. Nature 456, 66—72; 2008), she presented data on a second patient—tumour pair and preliminary data on a third. With hundreds of potential mutations churned up everywhere in the genome, her group focused on validating three different 'tiers' of single-nucleotide mutations, many of which lie in coding regions.
Asked why non-coding elements hadn't got more attention, she replied that her group was looking at these regions but that they would need more work to sort out, hopefully with the help of expression data and comparison with other patients. "Right now, it's not worth it," she said.
Nevertheless, Mardis remains a big fan of sequencing whole genomes. She says that the exome approach, which uses new techniques to capture the targeted DNA for sequencing, can miss as much as 20% of the coding regions.
"If the amount of data is scary, why not sequence the whole genome and then just focus on the genes?" she asks. "You could posit that is ultimately a cheaper approach than trying to get 100% of the genes, only coming up with 80%, and then going to some extraordinary measures to get the remainder that you missed."
Francis Collins, former head of the US National Human Genome Research Institute (NHGRI) in Bethesda, Maryland, agrees. "None of the methods are perfect," he says. But he predicts that in the near future, "exome sequencing is where most of the action is going to be".
And many cancer researchers see exome sequencing as a reasonable stop-gap solution until sequencing whole genomes becomes cheap enough. Lynda Chin of the Dana-Farber Cancer Institute in Boston, Massachusetts, says that exomes are a faster way in to identifying driver genes, and help accelerate better screening methods and treatments.
Chin has headed up some of the projects for the Cancer Genome Atlas (TCGA), a potentially billion-dollar-plus programme announced in 2005, and directed by the NHGRI and the National Cancer Institute. TCGA is now moving out of its pilot phase, in which it sequenced and characterized hundreds of tumours from three different types of cancer found in lungs, ovaries and brain, towards characterizing 20—25 cancer types.
In conjunction with some whole-genome sequencing, says Chin, exome sequencing will be part of the new portfolio. "We have to push the envelope," she says, "now".
Fascination bump. The body is so efficient it is hard to believe that the non- coding regions are not used for something. Tumors with multiple mutations, wow, not just one stray gene. I like the driver/ passenger mutation explanation. I have a teenage driver this summer, he will no longer be a passenger mutation.
My sister's sarcoma was a tumor composed of four distinct cancers.
How did they distinguish there were four? I suppose they looked very different under a microscope, my cytology is very rusty.
She had a recurrence and underwent proton radiation this spring.
I'm just praying for as much time as the good Lord will give her.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.