SAVING GRAPES
The most comprehensive genetic analysis to date of domesticated grapes reveals that they are a close-knit bunch, which may help explain their vulnerability to disease and point the way toward breeding hardier vines. While at Cornell's Institute for Genomic Diversity, Sean Myles, now a postdoctoral researcher in genetics at the School of Medicine, scanned the DNA of more than 1,000 leaf tissue samples from the U.S. Department of Agriculture's germplasm bank using a custom gene array.
He found that of the 583 unique grape varieties in the USDA's collection, nearly three quarters have a parent-child or sibling relationship to at least one other type. "Previously people thought there were several different families of grape," Myles told the New York Times. "Now we've found that all those families are interconnected and in essence there's just one large family." For example, Viognier and Syrah, two of the most common varietals grown in France's Rhône Valley, are likely brother and sister.
And while modern grapes have retained most of the genetic diversity of their wild ancestor, limited cross-breeding has left the plants vulnerable to disease. Grapevines are typically propagated from cuttings to ensure that they are genetic clones of the parent plant. As a result, their genes have not been shuffled much in the millennia since grapes were first cultivated, giving the microbes that prey on them ample time to gain the upper hand.
This research provides a first step toward identifying genetic markers for disease resistance that can be employed as part of "marker-assisted" breeding programs. "If growers are brave enough to discard 1,000-year-old cultivars like Pinot Noir, which are sitting ducks for pathogen pressures, and begin growing disease-resistant cultivars, we can drastically reduce the amount of chemicals we spray onto the world's vineyards," Myles said in an interview with the School of Medicine's Scope blog.
IMPROVING COMPUTER VISION
In January, Intel announced plans to invest $100 million over the next five years in academic research in areas of "vital importance" for the computing industry, including visual computing, mobility, security and embedded solutions. The microchip maker plans to establish centers at half a dozen universities by 2012, the first of which will be based at Stanford.
Led by professor of computer science Pat Hanrahan, the Intel Science and Technology Center for Visual Computing will focus on improving the way users interact with computers. "Visual computing is in a renaissance right now," Hanrahan said in a video news release, pointing to devices such as smartphones, iPads, large displays and enhanced game machines, each offering new modes of interacting with content. "The center is really organized to address research problems that those systems pose."
The Stanford center will receive $2.5 million a year for five years. More than 30 researchers at eight universities—UC-Berkeley, UC-Davis, UC-Irvine, Cornell, Harvard, Princeton and the University of Washington—will collaborate on projects ranging from user-friendly software tools for creating and controlling visual media to computer vision systems that can infer functional and semantic information.
