San Diego, CA, March 15, 2006 -- Three years ago, Ben Raphael knew very little about cancer biology. Then a team of cancer researchers approached his research supervisor at UCSD, Computer Science and Engineering Professor Pavel Pevzner, about developing new computational techniques for understanding the rearrangement of genes in tumors. "When working in bioinformatics, understanding the biology is a huge advantage," recalls Raphael, a fourth-year postdoctoral researcher in the Jacobs School of Engineering who earned his Ph.D. in mathematics from UCSD. "I work in a field that requires collaboration between biologists and computer scientists, and it helps to have knowledge of both disciplines."
"Our team at UCSF Cancer Center sequences tumor genomes to understand mutations in DNA that may scramble single letters, or manifest in much larger-scale rearrangements of whole blocks of letters such as chromosomal inversions, translocations, deletions and duplications," says UCSF's Volik. "We began to realize that the tumor genome is much more shuffled than anticipated, so we sought out new computational techniques to understand how a cancer genome breaks apart and gets put back together."
In other words, adds Raphael, the survival of the fittest also applies to tumors. "Cancer itself is an evolutionary process," he says. "The cells in a tumor are mutating and competing for resources, and the cell with the greatest advantage wins and begins to dominate the population."
Since joining the UCSF-led project in 2003, Raphael and Pevzner have developed algorithms for analyzing genome rearrangements in tumors using a technique called End Sequence Profiling (ESP). The technique builds on the completion of the human genome sequence. "Basically, if you want to find large-scale rearrangements of the genome, it is not necessary to sequence every letter of DNA," explains UCSF's Volik. "Since tumor genomes are rearranged and mutated versions of the human genome, we can sequence fewer base pairs of DNA and get information about these rearrangements. After sequencing short 'tag' sequences, we can see where they map to the human genome."
The March Genome Research paper serves as a demonstration of the power of ESP. Based on sequence data supplied by the UCSF researchers, Raphael ran his algorithms for further insight. "The computations suggested regions of the genome that should be examined more closely," he says. "The UCSF group then performed additional targeted sequencing of the tumor genome, and to a large extent the results supported the computational predictions."
In several types of cancer, specific rearrangements affect particular genes, and there are now two drugs that are targeted to specific genes that are rearranged in tumors. "When genome sequencing is much cheaper and we understand the structure of tumor genomes better, you can imagine that your doctor could catalogue all mutations in your cancer and create a cocktail of drugs that would counteract those mutations," says Raphael. "This project is a very small step towards that goal."
So far, Raphael and Pevzner have performed their computations on UCSF's sequence data for one specific breast cancer cell line (MCF-7). However, they are currently analyzing four additional breast cancer samples, as well as tumors from prostate, ovary and brain tumors. They are also developing algorithms for combining data from ESP with results from other experimental techniques used in tumor analysis. "Combining data from multiple sources will give us a lot more information at substantially less cost," marvels Raphael. "For the same cost as sequencing one human-sized genome, we will be able to map at least 100 different tumor genomes."
Widespread adoption of the technique could speed development of a cancer genome atlas - proposed in December by the National Institutes of Health -- based on the sequencing of a large variety of tumor types.
In the meantime, Raphael is nearing the end of four years as a postdoctoral researcher. "I extended my postdoc for an extra year because I received additional funding for this research," he notes, referring to a $500,000, five-year Career Award at the Scientific Interface (CASI), awarded by the Burroughs Wellcome Fund in December 2004.
Raphael was one of only eleven scientists nationwide selected for the honor.
The last three years of the CASI funding are designed to support young faculty researchers, so Raphael is now juggling job interviews across the country. He recently completed a swing that took him across the Midwest and East Coast. Next up: a tour of West Coast institutions, ending back at UCSD.
Genome Research »
UCSD Bioinformatics Researcher Studies Tumor Genome Architectures with Care »
Burroughs Wellcome Fund »
Ben Raphael Website »
More Bioengineering News Via RSS More Computer Science and Engineering News Via RSS
Get the monthly newsletter from the Jacobs School of Engineering