Researchers at the St. Jude Children's Research Hospital had their sights set on understanding the genetic basis of pediatric cancer long before they partnered with their collaborators at Washington University in St. Louis, and began to analyze hundreds of genomes.
"Many of the tumor groups have been pursuing pre-next-gen analysis for many years now ... and have made important insights, but have not been able to fully elucidate all the genetic changes that underlie childhood cancer development, or the ways in which they do or don't respond to treatment," St. Jude's Charles Mullighan says. "There were strong reasons to go ahead with next-gen sequencing."
The St. Jude Children's Research Hospital-Washington University Pediatric Cancer Genome Project, or PCGP, launched in January 2010, and its members aim to sequence 600 pediatric tumors and matched, non-tumor DNA — 1,200 genomes total — within three years. From the start, the team set out to achieve its sequencing goals in a tiered fashion, keeping pace with technological advances, Mullighan says.
"The first year was 50 tumors, which was completed. The second was 200, and we've exceeded that, [with] the balance [to be completed] in the last year of the project," he says. "We've now completed analysis on over 250 — and that's polished. We have raw data back for 300."
Among those 250 childhood cancers the team has analyzed are three that were recently published in two separate Nature papers and a Nature Genetics article, all of which appeared online in January. Members of the PCGP sequenced the complete normal and cancer genomes of four patients with retinoblastoma — identifying RB1 as a potential therapeutic target — and 12 patients with early T-cell precursor ALL — uncovering mutations and a gene expression profile that suggested ETP-ALL patients might benefit from treatment developed for individuals with acute myeloid leukemia.
PCGP members also uncovered links to mutations in genes involved in DNA organization in the diffuse intrinsic pontine gliomas they sequenced.
Post-publication, the St. Jude-WashU team has posted its genomic datasets on its Explore website, which was launched in January. "We have always felt very strongly that the genomic data needs to be deposited in the public domain, even before the journals were insisting on this," Mullighan says.
Of course, uploading whole-genome data to the Web is not a trivial undertaking, nor does it ensure utility, he adds. That's because whole-genome datasets are not very tractable to the average bench researcher, he says. "I can say from experience," Mullighan adds. "I used to download and analyzed gene expression and SNP array data myself, [but] I would never even try to analyze whole-genome data." As such, Explore includes searchable features — like the genomic alterations summary tool — that allow users to query the whole-genome data in targeted ways.
Mullighan says some PCGP members have found such tools useful. "We might be focused on our individual diseases as PIs, but our informaticist is saying 'You might want to look at this gene; there's only a single case that it's mutated here, but if we look across the project, other tumor types have the same mutation as well,'" he adds.