Washington

Researchers in genetics trying to share their results face an uphill struggle. The wide range of experimental methods that they use makes it fiendishly hard to compare data. But a consortium of companies and academics is now working with the US government to tackle the problem by drawing up a better standards system.

The External RNA Controls Consortium (ERCC) says that such standards will allow geneticists to reach more accurate conclusions, and to compare their data with the findings of others.

Sharing nature: fresh standards should make it easier to compare results from microarrays. Credit: J. CHENG

“There is definitely a need for standardization,” says Gavin Sherlock, a geneticist at Stanford University in California. “It's one of the biggest problems we have — there is no way of connecting the dots between two platforms very well.”

The bulk of genetics research involves characterizing the genes expressed by a particular cell or tissue under specific experimental conditions. But there are many ways to do this, and researchers say that they can get different results from the same experiment depending on the method they use. They can even have problems duplicating their own results from the same technology.

The ERCC is addressing the issue by setting up standard controls for checking the performance of experiments. The group is drawing up an extensive list of genetic sequences that won't normally turn up in most studies. Pieces of genetic material that correspond to these sequences will then be manufactured for use as controls, allowing geneticists to test how well they are detected in their laboratories. Researchers will then be able to compare the results of different experiments with more confidence, consortium members say.

The ERCC held two workshops last year, and gene-chip manufacturers such as Affymetrix, based in Santa Clara, California, are already gearing up to produce the materials that the plan requires. Marc Salit of the National Institute of Standards and Technology, which helped to organize the workshops, says that a non-profit means will probably be established to distribute the controls. He says that in the most optimistic scenario, the materials could be available within a year to 18 months.

The ERCC was set up in the wake of a standardization effort led by the Microarray Gene Expression Data Society, a group of scientists that has drawn up a common language for comparing gene-expression experiments and which is now devising data formats to help researchers to compare their results (see Nature 415, 946; 200210.1038/415946b ).

Janet Warrington, vice-president of Affymetrix, says that one of the ERCC's goals is to move genetics closer to the clinic. Making sophisticated genetic analysis accessible to physicians will require much smoother standardization than now exists, she says: “If you're going to develop something for clinical application, you really want to look into developing standards.”

Researchers say that the ERCC's work will be a big help, but will not entirely solve the problem of standardization. For instance, says John Quackenbush, a bioinformatician at The Institute for Genomic Research in Rockville, Maryland, the consortium's efforts are primarily focused on experiments in humans, so research in other organisms could be left out in the cold. “Will it solve all the problems? Absolutely not,” Quackenbush says. “But it's a very important first step.”