NEW YORK (GenomeWeb) – As CRISPR/Cas9 technology jumps out from academic research into the biotech industry, two startups are positioned to provide paying customers, in addition to academic researchers, with tools to manage their genome editing experiments.
Desktop Genetics, a spin out from the University of Cambridge, and Benchling, which was incubated at Y Combinator, are both poised to take their CRISPR/Cas9-related software platforms to organizations looking for software that can do more than just design guide RNAs (gRNA). And both are providing updates to their software to keep pace with the rapid development of the field. Last month, Desktop Genetics launched its DeskGen platform at SynBioBeta in London and last week Benchling announced a faster algorithm to calculate off-target effects.
Designing the right guide is the crux of CRISPR/Cas9 genome editing. The right guide will bind the Cas9 protein to the genomic region of interest — it's what makes the technology programmable— to get efficient activity; however, it must also not bind elsewhere and lead CRISPR/Cas9 into its infamous off-target effects.
With more than half a dozen tools available from academic groups, scientists have a multitude of options to turn to for help in designing the best gRNAs for their genomic target of interest. Desktop Genetics and Benchling, have joined in to offer free design tools with the hopes of showing off their other features for genome engineering experiments.
Both companies most apparent feature is their respective guide RNA (gRNA) design tools. As a class, the gRNA design programs can do everything from help find target sites to predict off-target effects. Some can help design gRNAs for as many as 10 organisms, while others focus just on human and mouse. Several were developed in pioneering CRISPR/Cas9 research laboratories, such as E-CRISP, Chopchop, COSMID, and the Feng Zhang lab CRISPR Design Tool.
At last week's Genome Engineering Workshop 3.0 at the Broad Institute in Cambridge, Massachusetts, Ann Ran of the Zhang Lab demonstrated the basics of gRNA design using gUIDEbook, a collaboration between Horizon Discovery and Desktop Genetics. She told GenomeWeb that her decision was not an endorsement of the tool over any other tool.
In comparison to earlier gRNA tools, gUIDEbook and Benchling offer slick interfaces and advanced features such as predicting off-target effects and designing pairs of gRNAS for editing with the "paired nickase" method of CRISPR/Cas9. The algorithms incorporated to predict on- and off-target activity of gRNAs and those to rank them are all based on the published scientific literature, Desktop Genetics COO Victor Dillard told GenomeWeb.
Using the gUIDEbook tool, researchers can browse for genes of interest or manually enter any sequence to knock in or knock out, and the tool will spit out a list of suggested gRNAs. Potential gRNAs are ranked based on how well the gRNA will bind to the target, as well as off-target activity, based on whether the off-target sites are in non-coding regions (good) or coding regions (bad), including tumor suppressor genes (especially bad). A visual display shows the guides as well as their binding scores and cut sites.
"Designing a good CRISPR [gRNA] is a balance of factors," Dillard said. "We saw that people used to use a variety of different tools. Any one scientist had their own method of designing a good CRISPR experiment," he said. That would often include multiple applications of commercial and academic software.
"What we tried to do was put all of that under one roof and integrate that," said Dillard. Rather than simply aggregate results calculated with a variety of methods, the firm chose the approach of evaluating the scientific literature for what it thought were the best rules for designing guide RNAs.
Binding scores in gUIDEbook, which calculate on-target efficiency or how well the gRNAs will stick to the target site, are available in the free "Wizard" mode and use the scoring system developed by John Doench of the Broad Institute. There's an "Advanced" mode that requires registration, but it is also free and provides off-target specificity scores throughout the entire genome, calculated using a method developed by Patrick Hsu and colleagues at the Broad Institute's Zhang lab. However, the software is modular, Dillard said, so any custom algorithm could be used to rank gRNAs. He added that users are split "about fifty-fifty" between the Wizard and Advanced design tools.
DeskGen's additional features can place the gRNA into any CRISPR/Cas9 vector as well as generate an assembly strategy for cloning. The guides are available for order from Horizon Discovery.
The web-based DeskGen platform only uses human and mouse reference genomes, but custom deployments of the platform can incorporate any genome, including genomes for specific cell lines that might be used to study disease, Dillard said.
Another web-based experiment platform, Benchling, offers a gRNA design tool alongside other DNA design tools. Benchling CEO Saji Wickramasekara said the firm last week launched a newer and faster version of its free CRISPR tool, which offers scores for ranking potential gRNAs by both efficiency (on-target) and specificity (off-target).
"Previously, it would take five or six seconds to calculate scores for each guide RNA," Wickramasekara said. "For hundreds of guides, that took a long time. Now we can calculate scores for 300 guides in three to five seconds."
As with DeskGen, Benchling's scores for efficiency and specificity are based on the same math used in Doench's and Hsu's methods, respectively.
Benchling's and Desktop Genetics' gRNA design tools are also selling points for other features. Desktop Genetics offers deployments of its platform that aren't web-based, with additional features such as loading any genome of interest. Dillard declined to name any specific customers, but said they include drug developers and reagent suppliers. Wickramasekara said Benchling offers features to address the compliance and security needs of larger organizations in its enterprise package.
It's not just scientists who are taking note of these software platforms. Benchling recently raised $5 million in private financing led by Silicon Valley venture capital firm Andreessen Horowitz. Dillard said Desktop Genetics is in the process of fundraising and expects to announce the results soon.