Pursuing personalized medicine through bioinformatics
These are the promises that have been ushered in by the post-genomic era, thanks to the sequencing of the human genome in 2000. The realization of these promises will depend in a significant way on the relatively young science of bioinformatics and the ability of bioinformaticians to leverage the ever-increasing power of supercomputers.
At the UB Center of Excellence in Bioinformatics, Director Jeffrey Skolnick, UB Distinguished Professor, and his team of researchers conduct their work on the world's largest supercomputing cluster devoted to bioinformatics. "Ultimately, we're trying to reduce the lead time for drug development, not for one molecule or a few, but for hundreds or even thousands," Skolnick explains.
Research at the center focuses on predicting the structure of proteins of a certain size, smaller than a few hundred amino acids, and figuring out how they fit into the amazingly complex biology of normal function and disease.
"Even if each simulation of each protein took five days of computer time, and even if you're looking at just 14,000 protein structures (the small proteins in the human genome), then you're talking about 70,000 days [or more than 190 years] to predict protein structures on a single computer processor," says Skolnick.
"So if you don't have on the order of several thousand processors, you just can't reasonably do it. But if you have 4,000 processors, which we do, then, theoretically, it could take only 18 days."
And that's only for one genome. Skolnick and his colleagues are studying numerous genomes, including those of pathogens and organisms such as the mouse, the basis for many models of human disease. They also study less complex genomes, an understanding of which will provide the basis for better approaches to the human genome.
After important protein structures are deciphered comes what Skolnick calls "the hard part": the analysis of each structure to determine its role in biochemical function and how it may influence, and be influenced by, cellular processes involved in disease.
"Ultimately, what we want to do is relate genotypewhat's happening geneticallyto phenotypewhat's happening clinically," he says. "We want to find out the physiological manifestation of this protein structure in this cellular pathway.
"The overall goal," he adds, "is to develop personalized medicine, which is based on understanding how a drug affects you, versus how it affects me."
Skolnick and his team already are making headway in connecting the dots. In November, the team published results that provide the first genome-based ability to predict protein-protein interactions, work developed while Skolnick was at the Donald Danforth Plant Science Center in St. Louis, Missouri.
"We are now moving toward an understanding of how the whole system works what's known as systems biologywhich is the key revolution in the post-genomic era," he explains. According to Skolnick, the Protein Data Bank, the international "public library" of solved protein structures from which scientists draw data, contains not just isolated molecules, but in many instances solved compounds consisting of two or more proteins interacting.
"Lots of cellular signals are mediated by these protein-protein interactions," he says, "but it's a very crowded party and we want to know exactly who's interacting with whom. Often, the function of one protein can be deduced by studying the proteins with which it interacts."
Skolnick conjectures that perhaps there are hundreds of millions of these interactions, a seemingly intractable problem. But, he says, the process of learning which proteins are interacting is accelerated greatly if researchers have a computational method to help pinpoint the sites on the interacting proteins that will help scientists discover their role in biochemical pathways.
"That's what our method aims to do," he notes. "Using our supercomputer, we can start to see how the paths fit together, how this enzyme interacts with that small molecule or functions in a cascade of cellular processes.
"Ultimately, you want to know how the expression of these particular molecules relates to a particular phenotype. For example, how a set of proteins causes a particular kind of cancer," he says.
For bioinformaticians like Skolnick, these unprecedented challenges carry with them unprecedented opportunity, as these scientists are getting a first glimpse of the dazzling array of complex interactions of the biochemical activities that make up human life.
"There is an immense and voracious appetite out there for the kind of data we are generating," says Skolnick.
return to High-speed science main page