Mention medical research and images of lab coats, microscopes, and test tubes spring to mind. Yet one vital tool fueling advances in precision medicine may escape our thoughts: the computer. Ben Berman, PhD, co-director of Cedars-Sinai’s Center for Bioinformatics and Functional Genomics, gives five reasons why computers should be added to our collective biomedical imagination.
1 The cloud’s the limit.
Geneticists rely on bioinformatics specialists like Berman because of the vast volumes of information through which they wade. DNA molecules are made up of chemical units in pairs, and every human genome has about 3 billion base pairs. Plus, new genomic technologies allow researchers to investigate changes between thousands of cells within a tissue, or across thousands of individuals. “It’s rapidly become information you cannot analyze on your desktop computer,” says Berman, who speeds up the process by running different sections of the genome across hundreds of computer servers at once. His programmers are experimenting with using commoditized cloud-based services. “Instead of spreading computations across 120 servers at Cedars-Sinai, by using a cloud, we could spread it across thousands of servers,” he explains.
2 The puzzle has many pieces.
Genes are not the final arbiters of our medical fates. Lifestyle, exposure to toxins, nutrition, and treatments all matter. Many health factors are summed up in the clinical data of our medical records. Combining that data with genomic information is key to unlocking a new era of precision medicine. “Going forward, genomic studies will be increasingly linked to clinical trials so we can see how the genome determines responses to certain drugs,” Berman says.
3 “Junk” DNA is littered with gems.
For decades after they first traced DNA’s double helix, many genetic scientists estimated that only the 2 percent of the genome that codes for proteins was worth a look. The rest was “dark matter” at best, and “junk DNA” at worst. After the human genome was sequenced — a colossal effort to which Berman contributed — systematic comparisons of different individuals’ genetic makeups revealed that this supposed detritus actually held many regions that act as switches, turning genes on and off. “Genomic regions that used to be considered junk encompass more than half the regions implicated in increased cancer risk,” Berman says. “This whole field of gene regulation — epigenetics — is very important.”
4 Software is king.
The challenge to bioinformatics is binding together an exponentially expanding universe of data to make it usable. Innovative software programs are starting to strategically connect clinical and genomic data. “In the research community, there is an emphasis on open software standards, and this paradigm should be embraced to link up medical organizations around the country,” Berman notes. “For precision medicine, this is a major task that involves technical, ethical, and regulatory hurdles. It has to bring together data scientists who work in diverse areas, including clinical information systems, mobile devices, and social media.”
5 Genetic analysis has gone global.
Berman is involved in the biggest effort ever to systematically analyze cancerous tumors. The International Cancer Genome Consortium is examining complete tumor genomes, including the non-protein coding regions, from 2,800 patients. This massive effort includes patients and data from more than a dozen countries, being analyzed by a consortium of more than 700 scientists (Berman co-leads the group evaluating epigenetic data). “Instead of moving patients’ genetic data to our hardware, we move our software to the cloud,” Berman says. The move protects patient privacy — a Parisian patient’s information will never be downloaded to a computer at Cedars-Sinai, for example — but enables experts from Los Angeles, Singapore, and France to jointly access information and work together to figure out what makes cancer tick.