New algorithm brings predictive modeling to gene research
Two University of Southern California scientists have created an algorithm to bring gene sequencing closer to clinical use.
The declining cost of gene sequencing, paired with advances in computing power have raised the possibility of more widespread use of genomics to provide personalized medicine. So far, though, cost has been a huge barrier.
The new algorithm, described in the journal Nature Methods, helps researchers determine how much DNA to sequence--enough to get the answer, but not too much that would waste time and money.
As data sets have grown exponentially, new thinking has been required about the mathematical properties of DNA sequencing data, the authors say in an announcement. Their method allows researchers to predict how much can be learned in a large-scale DNA sequencing experiment.
"It seems likely that some clinical applications of DNA sequencing will become routine in the next five to 10 years," Andrew Smith, a computational biologist at the USC Dornsife College of Letters, Arts and Sciences, said in the announcement. "For example, diagnostic sequencing to understand the properties of a tumor will be much more effective if the right mathematical methods are in place."
Bioinformatics and biostatistics are becoming ever more integral to DNA sequencing. The Association for Molecular Pathology reported in the November issue of the Journal of Molecular Diagnostics that the massive amounts of data produced threaten to create bottlenecks without adequate bioinformatic infrastructure.
Meanwhile, researchers at the University of Massachusetts-Amherst have applied sophisticated statistical tools to large public databases to ferret out clusters of gene variations in people with specific conditions such as heart disease.
Those variations, however, could raise concerns, but ultimately not cause any disease, which could lead to overtreatment, just one of the barriers to mainstream genetic tests outlined in a Wall Street Journal article last May. Still, the cost of a full genetic map is expected to fall quickly and tests are expected to increasingly be automated.