HONOLULU (GenomeWeb News) – Cancer genetics and genomics were hot topics at this year's American Society of Human Genetics annual meeting, with sessions focusing on everything from cancer genome sequence and structure to the genetics of cancer treatment response.
On the large-scale cancer genome sequencing front, Elaine Mardis, co-director of Washington University's Genome Sequencing Center, described her team's efforts to sequence whole genomes from individuals with acute myeloid leukemia and breast cancer. Mardis and her co-workers published the first cancer genome — from acute myeloid leukemia — last November and the second AML genome this August.
Mardis said the team is continuing to sequence and analyze AML genomes and also has started working on breast cancer samples. So far they've sequenced four breast cancer-related genomes: a primary tumor, a metastatic brain tumor, matched normal tissue, and a mouse xenograft tumor.
The team used an Illumina Genome Analyzer IIx to get 75 to 100 base paired-end reads. In so doing, they got between about 30 and 40 times coverage of each of the genomes, which they are in the process of analyzing. Mardis said they have already identified some large insertions, deletions, and other structural variants that appear to differ between the primary and metastatic tumors.
Meanwhile, Gaddy Getz described ongoing cancer sequencing efforts at the Broad Institute. Getz, a computational biologist and leader of cancer genomics analysis for the Broad's Genome Biology Program, said the team is doing whole genome, exome, and transcriptome sequencing for various cancer types as part of The Cancer Genome Atlas project.
Whereas the team sequenced one tumor-normal genome last year, Getz noted, they have now sequenced the whole genomes for 26 tumor-normal samples representing a host of cancer types, including multiple myeloma, prostate cancer, glioblastoma, and ovarian cancer. Their analysis so far has turned up mutations, copy number changes, and rearrangement patterns that vary by tumor type.
"The emerging picture from this research is that tumor genomes vary in terms of their rates of mutation and — even more pronounced — in terms of the number of rearrangements and copy number alterations that they harbor," Getz said in a statement.
Within the next year or so, he predicts that the team will have sequenced more than 100 additional matched tumor-normal pairs, with the goal of eventually generating whole genome sequence for around 500 such samples.
Getz said he and his colleagues also are working on transcriptome analysis and whole or partial exome sequencing for far more samples.
Both Mardis and Getz noted that while sequencing costs are decreasing rapidly, the expense of the bioinformatics used for storing, analyzing, and interpreting the sequence data is quickly becoming one of the most costly aspects of such research.
"The informatics cost — or just storing the data and analyzing the data — is growing compared with sequencing costs," Getz told reporters during a press briefing.