Bioinformatics software vendors have been reluctant to jump aboard the RNA interference bandwagon, perhaps having learned a bit from prior history. After all, researchers can access any number of free programs for designing short-interfering RNAs with just a Google search. And these programs don’t just come from academics, as many oligo vendors are happy to offer their design tools as a value-added service for potential buyers.
But as the RNAi market heats up, some vendors are turning to bioinformatics as a key differentiator, touting the ability of their algorithms to design siRNAs with a higher chance of silencing a gene of interest. Last week, for example, MWG Biotech added a free siRNA design tool to its e-commerce portal “to ensure design success” before users order their siRNA with “one simple mouse click.” Likewise, RNAi heavy-hitter Dharmacon upgraded the free version of the siRNA design algorithm it offers through its website — a pared-down version of the proprietary SmartSelection algorithm it uses for its siRNA synthesis service.
Meanwhile, bioinformatics startup Ocimum is taking the plunge with what appears to be the first commercial standalone siRNA design package. The company, headquartered in Hyderabad, India, with US offices in Indianapolis, expects to launch the product, called RNAiwiz, by March, and has estimated the market for the tool to be in the range of $50 million. Ocimum CEO Anuradha Acharya said that despite the proliferation of free tools, the demand for a commercial RNAi package is very real. The difference between RNAiwiz and freeware, she said, is the level of integration and data management that the company plans to offer as part of its product. “If people really want to benefit from RNAi, they should really be able to manage the data along with other biological data that they have — not just get help with designing one particular thing,” she said. Other options may be free, “but they’re not integrated enough,” she added.
Whether the market is willing to pay for such a commercial tool remains to be seen, but judging from the flurry of development among those offering freeware, bioinformatics is highly valued in the field of RNAi — whether there’s a price tag associated with it or not.
According to Acarya, identifying RNAi as “the next thing that researchers would be using” was easy. The challenge was in getting a viable commercial software package on the market in a timely manner. Founded in 2000, the company currently offers five products, ranging from LIMS to sequence analysis tools to microarray analysis software, but Acharya admitted that the company came a bit too late to the bioinformatics party to really make an impact in these areas. Now, she said, Ocimum has the chance to nab first-mover advantage in a new domain area, but the window of opportunity is narrow. “The people who enter the market first definitely have the advantage” in bioinformatics, she said. Citing firms like Silicon Genetics and BioDiscovery, she noted that “companies who came in first to the microarray space definitely had a big advantage over a company like us, who entered the market after two or three years … and they also can charge a premium for that.”
Ocimum is counting on its broad offering of software tools to help sell the new RNAi package. The company uses a modular approach in developing each package in its product suite, providing plug-and-play components that allow users to move data from one package to the next in a seamless manner. Acharya said that RNAiwiz will simply be another module in this integrated system, which also interfaces with the company’s LIMS product, so that users can analyze their RNAi data in the context of their other genomic information. The package is built on a client-server architecture and uses a relational database to store siRNAs along with statistical analyses on failed and successful gene-silencing experiments.
Ocimum currently employs around 55 people in India and 10 in the US, but Acharya said the firm is in ”expansion mode,” with plans to add another 20 employees over the next few months. In addition, she said, the company plans to ramp up its marketing and sales efforts to gain more visibility in the marketplace as it prepares to roll out the new software.
Web-Based Window Shopping
For MWG of Ebersberg, Germany, which offers bioinformatics services but not shrink-wrapped software, the new siRNA design tool serves as the linchpin of its RNAi e-commerce strategy, luring in customers not only for siRNA synthesis services, but also for additional bioinformatics service contracts. “There’s a big difference between offering convenient tools online to encourage people to use your e-commerce site rather than someone else’s, and offering a more complex bioinformatics service,” said David Brett, bioinformatics group leader at MWG. According to Brett, the company offers both options when it comes to siRNA design, but only charges for the latter.
“These tools that you see online from all of our competitors — and ourselves included — they’re normally to do with one, two, three, four, or five genes. But when you get into the area of 100, 150, or 100,000, then you need some more intelligent bioinformatics in the background, and then you really have to make it a project, and then you have to pay for it.” Brett said that his team is currently engaged in such a project, designing for a client hairpin-loop siRNAs for 1,200 genes. “We think trying to do focused, smaller bioinformatics jobs with a concrete biological product in association is a far better model than trying to sell pure bioinformatics by itself,” he said.
The new online tool does offer some advantages over other freely available options, Brett noted: In particular, it is designed to minimizeoff-target effects and distinguish between alternative splice and isoforms of the target gene. The tool offers a Blast function that automatically checks every candidate siRNA to see which gene or alternative splice form it matches, as well as a Blast function for the sense strand to discourage off-target effects.
Brett said that because MWG entered the RNAi market a bit later than some of its competitors, “we tried to go one better in that we’ve increased the ability of the software to look at those off-target effects through a double-Blasting of the reverse complement.” But despite the improvements in the software, it’s still based on publicly available methods, which rules out a fee-based model, he said. “As a marketing or sales strategy, it’s very difficult to launch a tool where you charge a fee for service when the rest of the competition has a certain tool — whether it’s good or not — for free. So you’re a bit stuck as a firm to be able to charge for software unless there’s something dramatically secret in your IP.”
Staying on Top
Dharmacon, however, is counting on just that strategy to keep it on the top of the RNAi informatics heap. The company’s new, upgraded freeware uses eight distinct siRNA design criteria — up from the four used in the previous version — and its release coincides with the publication of a paper in Nature Biotechnology describing its algorithm for “rational” siRNA design. But Dharmacon isn’t giving away all of its secrets — the paper only discusses a fraction of the 25 criteria that Dharmacon uses in the proprietary version of its SmartSelection algorithm.
Stephen Scaringe, Dharmacon CSO, said that publishing the method serves as “proof of concept of rational design for siRNAs” — an approach that he said is still met with skepticism in some circles “The ability to rationally design siRNAs is important and valuable in using siRNAs,” he said, as is publishing results in a peer-reviewed journal “that is trusted by our customers and the people we work with.” The paper should “give confidence to people that not only our early algorithms, but also our later algorithms are very well thought through, researched, and developed,” he said.
The eight criteria discussed in the paper include low G/C content, low internal stability of the sense 3’ end (or antisense 5’ end), a lack of internal repeats, an A at positions 3 and 19, a U at position 10, the absence of G at position 13, and the absence of G or C at position 19. Scaringe said that these, plus 17 other criteria, are weighted in the company’s proprietary algorithm, to account for the fact that some parameters are more influential than others. He added that the company’s ongoing bioinformatics collaboration with Rosetta Inpharmatics has also led to a reduction in off-target effects.
But Scaringe said that although Dharmacon’s approach to selection is “by far the most advanced” in the field, the proof is in the pudding when it comes to keeping customers satisfied. “They aren’t interested in siRNA, what the sequence is, how the RISC complex works, and all that cool stuff — most of them have a gene, and they want to know, ‘What happens if I silence it and that gene is not functioning?’” he said. “They don’t care how we do it.”