I’ll be spending the next few days at the Biology of Genomes meeting at Cold Spring Harbor, NY – one of the most awaited events on the genomics calendar. I plan to blog here about the major themes emerging from the meeting; you can also follow me on Twitter if you want shorter, punchier updates, and I’ve set up a FriendFeed group for more complex topics.
The meeting kicked off last night with a session on cancer genomics that gave a sense of the serious amounts of data currently being generated on the genetic origins of tumour development. Most of the work in this area has a fairly homogeneous flavour: researchers take tumour and normal tissue from the same individual, then sequence either the whole genome or a substantial fraction of it and look for differences in the cancer genome not present in normal tissue. Such differences (termed somatic mutations) reflect genetic changes that have occurred during the evolution of cancer in that patient.
However only some of those changes will actually be causative mutations (i.e. actually responsible for the progression of orderly human tissue into the sprawling cellular anarchy of cancer), while others will simply be side-effects of the increasingly unstable genetic architecture of cancer cells. Distinguishing these two categories of mutations – called drivers and passengers – will require sequencing very large numbers of cancer samples to identify those changes that occur multiple times in independent patients.
The ideal experimental strategy, then, is to sequence the entire tumour and normal genome for thousands of individual patients. That’s a daunting task, but advances in DNA sequencing technology are making it feasible to consider projects of this scale.
Last night we got a sense of the findings emerging from large-scale analyses of cancer genomes.
Mike Stratton from the Sanger Institute presented an overview of structural variants (large insertions, deletions, and other rearrangements of DNA) present in 24 cancer samples – the sheer number and diversity of the observed changes was overwhelming. Other speakers (e.g. Elaine Mardis from Washington University) presented high-resolution snapshots of individual cancer genomes generated from whole genome sequencing, giving a sense of the challenges involved in identifying and validating the variants involved. Some of the patterns emerging from these studies were hinted at by previous small-scale studies, but our new-found ability to explore diversity over entire genomes is allowing researchers to identify completely unexpected findings.
The overall picture was pretty clear: large-scale DNA sequencing is already transforming our understanding of the genetic architecture of cancer, but we are still at the very beginning of this process.