via genomeweb

By Ben Butkus

NEW YORK (GenomeWeb) – Fresh off a private financing round in which it raised more than $25 million, genomics firm Miroculus is looking to drastically simplify and reduce the cost of library preparation for next-generation sequencing with its fully automated, "sample in, library out" microfluidics platform.

The San Francisco-based company has developed a prototype system called the Miró Canvas, which leverages a core technology called Aeros that marries microfluidics with electromechanical force to manipulate fluids across a patterned electrode surface to perform a wide range of NGS library preparation processes.

Miroculus hopes to have an early-access version of its platform available in the summer and has already validated a protocol for whole-genome sequencing library prep with collaborators at the Broad Institute.

As it moves toward a potential full commercial launch at the end of this year, the company is developing a number of other library prep workflows for the platform, including for RNA sequencing and targeted sequencing, and plans to support customers who want to develop their own workflows, company executives said.

Miroculus emerged from stealth mode in 2016, focusing initially on microRNA-based diagnostic testing. Soon thereafter the company acquired Toronto-based Kapplex, a firm that had been working to commercialize a digital microfluidic platform developed at the University of Toronto. Miroculus originally aimed to use the platform to simplify sample prep for miRNA detection, but soon switched gears.

"While developing the technology to detect those molecules, we realized [it] was very powerful, and could [also] enable us to automate a number of different complex workflows for a number of applications," said Alejandro Domeyko, cofounder and CEO of Miroculus. "That's when we decided to start pursuing other applications such as library prep for sequencing."

The prototype Miró Canvas platform has three components: a benchtop instrument, a disposable microfluidic cartridge about the size of a well plate, and specialized software that allows users to create library prep protocols from scratch or edit existing protocols. For sequencing library prep, a user simply adds the necessary reagents and purified nucleic acid and "the system does the rest," said Adam Lowe, acting chief commercial officer and strategic advisor for Miroculus.

"[The software] is coupled with an algorithm that basically simulates whatever [library prep] protocol you want to work with, and how it would run in our system, so you don't have to think where you're going to mix, do PCR, or do cleanup," Lowe said. "It tells you that if you want to run this protocol you've created, then just load the sample and reagents in this inlet, and the protocol will be fully automated."

According to Domeyko and Lowe, Miroculus' platform will be competing mostly with a multitude of laboratory automation solutions because nothing like Miró Canvas exists on the market yet. The company has not yet provided a potential price for its platform, however.

"If you look at what people are using today, less sophisticated labs without a lot of resources in terms of capital budget and programming may be doing manual preps with multi-channel pipettors," Lowe said. "As you get to the medium and larger-sized labs, the vast majority of people are using more conventional lab automation platforms, like liquid handlers. The evolution on the sequencing side has been incredible, but the liquid handlers look pretty much the same as they did 20 years ago. There just hasn't been a lot of advancement or revolution there."

Lowe added that companies have been working to make sequencing more accessible to everyone by bringing down costs and by shrinking and automating sequencing instruments, but despite these advancements, NGS library prep remains a bottleneck. "That means that customers are probably making decisions about what they do based on ... the complexity of the workflows, and how much time and effort it takes to run those" he said. "We saw a path to making all those workflows easy and accessible on a very compact and portable platform."

What's more, by drastically reducing reaction volumes and simplifying the workflow, Miroculus thinks it can improve sequencing library quality and, in turn, sequencing data quality. Early results from beta testers of the platform seem to bear this out.
At the Advances in Genome Biology and Technology annual meeting in February, researchers from the Broad Institute and Miroculus presented results from a study validating Miroculus' Aeros technology for PCR-free WGS library prep and comparing it to the Broad's in-house workflow.

Maura Costello, a principal process development scientist at the Broad and first author on the AGBT poster, said that the Broad's current library prep workflow uses multiple commercial solutions.
"In terms of workflow ... at Broad, we're really good at doing thousands of samples in a sort of faster setting. We have a certain turnaround time, and methods that we do, but we're not good when you have to do one special sample that needs to be done faster – that's where everything sort of breaks down," she said. "Right now, our solution is to have a lab tech to go in the corner and do it by hand, and that's not really ideal. It's a version of the Kapa Prep Hyper prep protocol, but we've sort of automated it our own way over the years to get it to the scale we have it at currently – using liquid handlers wherever we can, connected to our homegrown LIMS system to do the tracking and chain of custody."

The main benefit of the Miroculus platform, she said, is that it enables reproducibility, tracking, and automation on a single-sample scale. "You just set it and forget it and walk away," she said.

She added that her lab was "pleasantly surprised" about the quality of the data the platform yielded.

Specifically, Costello and colleagues found that Aeros reproducibly generated library yields equivalent to manual benchtop experiments from about 100 ng to 200 ng of sheared input DNA (compared to more than 200 ng using the Broad's in-house method), and with about a third less reagent volume.

They also sequenced 19 PCR-free whole-genome sequencing libraries prepped by Aeros to various coverages and compared the results with 515 plate-based automated PCR-free libraries. The Aeros libraries achieved 95 percent coverage at 20X sequencing depth, with about 4.8 Gb less sequencing data. If sequencing were to be performed on an Illumina NovaSeq S4, that would translate to potential cost savings of about $50 per sample, the researchers noted.

In addition, the Aeros libraries achieved the same SNP sensitivity using about 16 Gb less sequencing data. Finally, on average, the library prep time was about 2.5 hours using Aeros and 3.5 hours using the Broad workflow, while the Aeros required two "user interventions" compared to the Broad's 12.

"I was assuming the results would be equivalent to what we do, but we were actually seeing improvements – hitting our genome coverage deliverable in fewer gigabases, which is great," she said. "And that was with reduced volume reactions, too. So it's basically saving money in multiple ways. If we can shrink our reaction volumes down, and stretch those library kits farther, then we can also maybe pool more samples together on a sequencing lane, because we can get coverage with [fewer] gigabases. Anywhere you can squeak out savings is a good thing for us."

Miroculus' Lowe said that the company hopes to have a product ready for the market by the end of this year, work that will be fueled by its $25.6 million equity financing, for which the firm filed a document with the US Securities and Exchange Commission in February. The financing has closed, Lowe noted, and the company has not disclosed its investors.

And while PCR-free library prep for WGS is the near-term focus, the company is also working on enabling library prep for RNA sequencing and targeted sequencing.

"Then behind that, we have a pipeline of different applications [for which] we're working on agreements with collaborators to develop," Lowe said. "We really want to come out with a full suite of applications that are supported, so if a customer doesn't want to develop their own protocol, we'll have a solution. And for those who want to port a protocol that we don't support onto the platform, we'll provide very intuitive tools for that. They don't need to know a programming language, they don't need to tell the system what reagents to move where — the system will figure it all out."

Longer term, the company envisions its technology supporting clinical applications. "The ability to go directly from a biological sample to a prepared library – that's sort of been missing in clinical workflows for NGS," Lowe said. "And if you think about being able to accelerate the process and make it decentralized, that's also really interesting in terms of the long-term trajectory."

Costello is already on board with that. "One idea that we were thinking about is: Can I take a blood sample, pop that in, do a DNA extraction and a library prep — so it's sort of a blood-in, ready-for-sequencing-out machine," she said. "That would be very transformative for clinical [work] and is something we'd really like to mess around with when we get our [beta] platform."