CRISPR Strategy & Standards: How Synthego’s Free Software Lays The Groundwork For CRISPR Futures

By Allison Proffitt

March 21, 2018 | “Science really moves forward when things become easier to the scientist: when you don’t have to spend as much time doing something, or a new capability comes on the scene,” says Paul Dabrowski, CEO and co-founder of Synthego.

CRISPR and gene editing are certainly doing their part to move science forward, but Dabrowski and his colleagues believe the real leap will come from making gene editing technology widely accessible and driven by standards.

In August of 2016 Synthego launched its flagship product: CRISPRevolution, synthetic guide RNAs for CRISPR/Cas9 genome editing. Early in 2017, the company announced a round of financing, particularly notable because it included personal investments from Jennifer Doudna and her husband, Jamie Cate.

Since then, Synthego has been growing. With a headcount of about 90 now, Dabrowski said the company is “feeling a little more substantial.” Last October, the company signed a distribution agreement with Thermo Fisher. The company has a gene knockout kit with a guaranteed editing rate of greater than 50% in any human cell line, and is working on a digital workflow that includes the design and analysis component of genome editing research that people do on computers. “We’ve continued to upgrade our design tool; it’s becoming more and more powerful,” Dabrowski said.

And in late January Synthego announced ICE, or Inference of CRISPR Edits, an open source algorithm for validating CRISPR edits that offers rapid, reproducible analysis of Sanger sequencing data. The company published a preprint outlining the tool on bioRxiv (doi: https://doi.org/10.1101/251082) and shared the source code on Github.

ICE Beginnings

“All of our products come from internal needs,” Dabrowski explained. “We’re up front trying to see how you do genome engineering as best as we possibly can. And we have our engineers developing tools side-by-side with the scientists.”

Scientists wanted a tool that would give high-quality, high throughput analysis results quickly using Sanger sequencing, Dabrowski said. Next-generation sequencing is fast and accurate if a lab has the right equipment, but researchers were looking for a solution for labs without ready access to next-generation sequencing. According to Dabrowski, there’s quite a bit of backlog in the NGS services world, especially in the Bay Area. “We’ve definitely had some serious delays in the past year or so,” he said.

TIDE—or tracking indels by decomposition—is a method that uses only Sanger data and can analyze a mixed population, but users must manually tune algorithm parameters and process each sample. TIDE is also limited to processing single guide experiments, and can’t handle experiments where multiple guides are used simultaneously. Finally, the tool does not scale well to many samples, writes the Synthego team in the ICE paper.

ICE was developed by working to improve TIDE by supporting more analysis cases and making the algorithm more robust. ICE offers batch upload, high-quality analysis of 100s of samples simultaneously, and generates publication-ready figures.

“It gives really high-quality results on larger edits. When people are removing large portions of their DNA with gene editing, there aren’t really other tools that can do this other than next-generation sequencing,” Dabrowski said. ICE lets anyone upload Sanger sequencing data and, “results come back in a few seconds.”

“In comparison to TIDE, ICE is able to analyze more types of experiments, requires no subjective parameter tuning, and has comparable results,” the Synthego team writes in the ICE paper. “These advantages results in a webtool and software package that is able to easily process hundreds of CRISPR editing results in a reproducible manner.”

The ICE algorithm includes a score called ICE-D that flags large or unexpected edits. “ICE-D provides a form of insurance for the ICE proposing process by being able to capture unexpected indels,” the authors write. “We have validated the robustness of the ICE tool by running analyses for thousands of Sanger files in one batch.”

Free Model

ICE and the Synthego design tool are both available for free at ice.synthego.com and design.synthego.com. It’s a business model to which Dabrowski is committed.

“We believe that the digital tools and the algorithms should be free to use, and we think the value that we bring to the table at Synthego is the simplification of going between these tools to the actual experiment,” he explained.

“Where you can see this going is, more and more of the workflow is digitally accessible. That’s our plan as a company: make these tools, make them easy to use, and give everyone confidence to use them. Then make sure our services fit in very nicely into this.”

Customers are happy. Daniel P. Dever, an instructor at Stanford University Medical Center in Matthew Porteus’s lab, has been working with several providers of synthetic guide RNAs, and has been pleased with Synthego.

“Synthego has value in this market because they are able to provide chemically modified sgRNAs in 7 business days and, importantly for our lab, at a small scale that is cheaper than other providers, so we are able to screen sgRNAs for activity in human primary cells. Historically, we would have to clone sgRNAs and test them in cancer cell lines. All in all, this business model saves us a lot of time, and time is money, and importantly we get data in primary cells quicker, which can push our translational projects quicker,” Dever wrote to Bio-IT World in an email.

He’s pleased with ICE as well. “ICE not only produces reliable indel data compared to the alternative algorithms, but it also generates publication-quality indel spectrums and Sanger traces instantly,” Dever said in a quote shared by Synthego. “ICE has the potential to become a standard for CRISPR analysis.”

Building a Full Stack

But all ICE users aren’t Synthego customers, Dabrowski said. There are ICE users doing plasmid-based or in vitro transcription-based editing, and the company has gotten in-bound requests from active and new customers that want private versions of ICE that can be deployed behind company firewalls. “It’s a service we’ve been looking at developing,” he said.

Synthego’s focus continues to be: “How do we continue to simplify genome engineering workflows,” Dabrowski said. The company is building toward, “what we consider a full stack for genome engineering, kind of like there’s a full stack in the computer world for programmers.”

Dabrowski declined to completely define that full stack, but said there are still functionalities in the CRISPR world that Synthego hasn’t yet tackled. “There’s still functionality in the CRISPR world both in terms of simplifying how the experimental workflow moves forward, and also in terms of the capabilities of the tools. Right now we’re really focused on knockouts… But there’s definitely interesting things in the works for this year,” he said.

Some of that simplification will involve training, Dabrowski said. CRISPR protocols still include a lot of variation and optimization, he said, and a standard framework for approaching those will be important. Training—even formally in a classroom setting—is something the company is thinking hard about, Dabrowski said. “How do we want to expand this? How far do we want to go?”

Dabrowski contends that there are, in fact, limited CRISPR users in biopharma because of the lack of standards. “There’s still a huge educational component. If there aren’t any standards, then it’s hard for people to really get behind it.”

Synthego sees training opportunities not only for staff using gene editing and the Synthego tools, but also at higher levels of the organization.

“Typically, whenever we’re engaged with someone at [the vice president or C-level of an organization], it’s a very quick conversation for them to understand that this has dramatic impacts on pretty much all aspects of R&D in their organization,” Dabrowski said.

He compared CRISPR to trends he’s observed in machine learning and artificial intelligence. “Pretty much every company has to have a strategy of how they make use of machine learning and AI for their data, for their business. In the biotech world, we’re seeing people start to wonder: What’s my CRISPR strategy? What’s my genome engineering strategy?”

The pace of R&D will speed up dramatically in the next few years, Dabrowski believes. Companies without a gene engineering strategy will “risk being left behind in big ways.” He predicts this year will be a tipping point—not to fully-defined CRISPR strategies, but the year companies begin prioritizing the conversation.

“Frankly, I don’t know that the field of CRISPR has the standards and clear direction forward yet. There’s some potential winners. And there’s some potential ways of simplifying these things. But it hasn’t fully played out,” he said.

Synthego, with what Dabrowski calls “state-of-the-art” CRISPR services, plans to have an important voice in conversations about the future of gene engineering. “If we’re successful as a company,” he said, “we may be a bit of a focal point of where genome engineering technology is going.”