How a stealth startup’s $100 genome could change how we treat and diagnose diseases like cancer

After six years in stealth mode, U.S.-Israeli startup Ultima emerged this summer with the announcement of $600 million in venture funding from General Atlantic, Andreessen Horowitz, Lightspeed, Khosla Ventures, the Founders Fund, and others—and the even more head-turning claim that its new device could sequence a whole human genome at a cost of $100, six times below what had previously been the cheapest option.

In 2001, sequencing one person’s whole genome—used as a way to find disease-causing anomalies in a person’s DNA—cost about $95 million. Since then, relentless technical innovation has steadily pushed the cost of sequencing down, leading to widespread adoption of the technology to understand human illness and predict health risks, to develop targeted drugs and diagnostic tests, and to identify infectious diseases and biological threats. 

In 2014, industry leader Illumina achieved a $1,000 genome with its HiSeq X Ten device. By 2020, its NovaSeq 6000, running full tilt, could sequence entire genomes for a cost of $600. With cheaper, widely available sequencing, the argument goes, we could do a lot more. People could get a routine annual screening for changes in DNA, from a blood sample, that would show signs of cancer far earlier than traditional tools. Biotechs could screen a billion—not a thousand—biological compounds at a time to find new treatments for rare diseases 

For years, one company, Illumina, with 2021 revenue of $4.5 billion, has enjoyed a near monopoly on sequencing, controlling about 90% of the market for machines, and compiling 80% of the world’s genomic information each year. But Ultima’s paradigm-shifting device has the potential to greatly reduce the cost of detecting illnesses like cancer—and disrupt Illimina’s industry domination. 

Ultima was founded in 2016 by Gilad Almogy, a veteran engineer at Applied Materials—a supplier of equipment, services, and software for manufacturing semiconductor chips, flat-panel displays, and solar panels—and the founder and CEO of Cogenra Solar, a developer of photovoltaic roof shingles, acquired by SunPower in 2015. Ultima’s UG 100 sequencer is inspired by the optical inspection systems used in those industries, says Ultima’s chief commercial officer, Josh Lauer. “It’s not too dissimilar a problem [to next-generation sequencing] because you’re using high-end optical systems, and you’re looking for one defect in, like, millions of transistors.”

The company’s large sequencing device is aimed at large academic and commercial, and government users doing large-scale, well-funded research projects. Now in early paid release with customers including MIT and Harvard’s Broad Institute of biomedical and genomic research center, the machine will be priced in line with Illumina’s high-end NovaSeq machines—$1 million dollars, give or take—says Lauer. But thanks to its unique architecture, the company claims it will be far faster and cheaper to run than Illumina’s devices. 

Both Ultima and Illumina use a similar approach to sequencing that involves cloning DNA and using miniature cameras to capture fluorescent signals indicating the chemical bases making up the strands. One big difference, though, is instead of using a specialized microfluidics chip called a flow cell—a glass slide etched with tiny channels and nano-sized wells that anchor the DNA, Ultima’s device uses a standard 200 mm semiconductor wafer—a circular disc. “It’s big and cheap,” says Lauer. “And it’s an offshoot of the global semiconductor supply chain, so it’s easy to source.” (The widespread semiconductor shortages affecting the electronics and auto industries haven’t impacted Ultima, which doesn’t require the same specially primed, “doped” discs needed for those applications.)

The ability to spin the disc also creates efficiencies in imaging, he says. Capturing images in a flow cell, says Lauer, “is kind of like mowing the yard. You go back and forth and back and forth, and the cameras have to slow down, change directions, and accelerate again.” Ultima’s radial imaging system works like a CD player: as the disc spins, the camera holds steady, following a continuous line without having to stop and turn around. That makes sequencing about twice as fast, says Lauer. 

One potential drawback to the device: The instrument is a single-end reader, meaning it sequences strands of DNA in one direction only, rather than front-to-back and back-to-front, which some experts say is less accurate at detecting certain kinds of genomic rearrangements and repetitive sequences that can be key indicators in cancer detection and treatment, for example. “Error rates are much higher than for Illumina,” says Rameen Beroukhim, an associate professor of medicine at the Dana-Farber Cancer Institute and Harvard Medical School, and an associate member of the Broad Institute. “In understanding genetic alterations—especially rare mutations—it’s important to get really accurate reads.”

Lauer says that Ultima’s “overall error rates are very similar to Illumina” and that the company is actively partnering with several leading oncology providers to show Ultima accurately captures clinically relevant mutations, and in some cases, even higher sensitivity, at lower cost.

But, even with reduced accuracy, cheaper sequencing could be helpful for applications like RNA sequencing and the detection of chemical changes that affect gene activity and are involved in many types of cancer, says Keith Robison, a computational biologist in the next-generation sequencing group at Boston-based Ginkgo Bioworks. A growing body of publications with Ultima’s research partners show “respectable results” with these types of analyses, says Robison. 

And cheap sequencing with less-than-perfect accuracy could be really useful in cutting-edge cancer diagnostics. “More sequencing can be applied in two ways,” says Robison. “To run more samples [at a time], or to get more data per sample, which we call depth.” Most tumor samples, particularly in colon cancer screening, contain non-tumor material that can create sequencing “noise.” Being able to cheaply run more tests could help find those needles in the haystack, he says. Indeed, Ultima has already announced a partnership with Exact Science, which markets Cologuard and other cancer screening tests. 

The challenge for Ultima, says Robison, comes down to “how many places will be convinced to buy it as their one instrument? And how many places have the volume of work to justify needing two instruments.” Says Lauer at Ultima: “We have 11 of the largest genome customers—including the New York Genome Center, Baylor College of Medicine, and the Sanger Institute—in our early-access program, and we’re focused on making these customers successful.” The company also announced a partnership with Regeneron Pharmaceuticals to speed drug discovery and development. 

“The original thesis of the company is that there’s a virtually unlimited need for genomic data, and when you create an architecture that can offer multiple orders of magnitude of cost reduction, scientists will find uses for it,” says Lauer. And they appear to have gotten Illumina’s attention. Soon after Ultima announced its device, Illumina, in late September announced its new sequencer, the NovaSeq X Plus, with an advertised cost-per-genome of $200.