When Dharmacon established the Genome-Wide RNAi Global Initiative late last year, one of the company's goals was to help establish standards for conducting RNAi screening experiments (see RNAi News, 10/7/2005).
Following the conclusion last month of the second meeting of the initiative's members a group of international non-profit biomedical research centers at appears Dharmacon is close to doing just that.
"Twelve members of the RNAi Global Initiative conducted the same experiments using the same protocols in the most extensive multi-site comparative RNAi screen conducted to date," Bill Marshall, vice president of technology and business development for Dharmacon parent Fisher Biosciences, said in a statement this week. "By comparing this large data set, we were able to identify several critical areas of potential variation and the key information required for inter-laboratory experimental comparison."
Marshall told RNAi News this week that the company is planning on publishing the findings and hopes to have a standards proposal available before the next meeting of the initiative, which will take place in Europe this fall.
"Twelve members of the RNAi Global Initiative conducted the same experiments using the same protocols in the most extensive multi-site comparative RNAi screen conducted to date. By comparing this large data set, we were able to identify several critical areas of potential variation and the key information required for inter-laboratory experimental comparison."
Marshall declined to comment in detail on the experiments conducted by the initiative members before they are published, but said that there were two different studies. "One was more rudimentary and the other one was more difficult, but you could essentially do them on the same cells and you could do them with the same [siRNA] collection," he said.
For the experiments, Dharmacon "distributed kinome, phosphatome, and cell-cycle [siRNA] sets about 800 or so different genes targeted to each of the groups, [which] did the exact same screening protocol and experiments at all the different sites around the world."
The company then "pulled all the data back into a central processing facility … at Dharmacon," he said. "We did a lot of number crunching, et cetera, to look at the comparability" of the data.
"We spiked in known positives, and we had a variety of different negative controls, et cetera," Marshall explained. "The things we were trying to look at were how many different negative controls should you have, and how [to] deal with plate position variations," for example.
"The results are quite fascinating," he said. "Despite the fact that it's the same protocol, it's the same collection, it's the same cells everything is identical you did see some variation, [which] then allowed us to identify those parameters that are going to be key to standardization."
Among the aspects of screening experiments found to be critical to standardization, Marshall said, were "the controls, as we all expected. The appropriate type of control [and] how many negative controls you do is going to be vitally important," as are positive controls and their positioning.
Also found to be key is how a researcher deals with "edge effects, [which are] dependent upon the assay itself … [and] how detailed you get in terms of the exact protocols," he said. "These are all key components that came out.
"Of course, a big part of this is upfront validation of the assays prior to putting them into the large screening protocols," he added. "So understanding the assay well, understanding the signal-to-noise ratio, and then understanding how it is you can minimize the potential for artifacts is key."
He said Dharmacon plans to publish the results of the screening experiments in an undisclosed peer-reviewed journal, as well as issue a proposal of experiment standards termed MIARE, or Minimal Information About RNAi Experiments for review and comment by the broader scientific community. Ideally, the company would be able to do both before the third meeting of the initiative planned for this fall in Europe, Marshall noted.
"It is our hope that we can do that," he said. "We can never predict the outcomes of the science, but we're working hard to get that done."
Additionally, Dharmacon is now working on "continuing to do this type of experiment with the [initiative] members [to show] that when you standardize around a certain set of parameters, you see excellent reproducibility," Marshall said.
Doug Macron ([email protected])