Gilad Almogy, PhD, always had a goal of producing a machine that would enable affordable, high-throughput sequencing when he founded Ultima Genomics eight years ago. The company has now reached the milestone of producing a machine that can sequence a genome for $100.
Almogy, CEO of the company, studied quantum optics and spent time developing semiconductor-based diagnostic technology before moving into the sequencing field. He has applied this knowledge to redesign high-throughput sequencing and create a completely new machine. Investors clearly like the new technology and have contributed over $600 million in funding to Ultima over the last few years. The company also has partnerships with big industry players such as Regeneron and Nvidia.
The company’s UG 100™ sequencer has dispensed with the flow cell set up that many other sequencing machines use. Instead, sequencing beads loaded with DNA are sprayed onto disc-like round silicon wafers. Reagents for the process are dispensed from above and the wafers spin, allowing quick distribution of all the ingredients needed for the process via centrifugal force. The wafers are also scanned from above during the process allowing a continuous ‘read’ to be collected and interpreted with the help of artificial intelligence.
Although the technology is very promising, Almogy and colleagues face stiff competition from industry giants Illumina, which released the high-throughput NovaSeq X range of machines at the end of 2022, and MGI Tech subsidiary Complete Genomics, which unveiled its high-throughput DNBSEQ-T20×2 machine early last year. However, Ultima hopes its high-throughput sequencing, affordability, and accuracy in single nucleotide polymorphism (SNP) calling will make its newly available machine attractive to the cancer diagnostics field. For example, testing for minimal residual disease in patients who appear to be in cancer remission.
Almogy spoke to Inside Precision Medicine senior editor Helen Albert about his entry into the world of sequencing, developments at Ultima, why the $100 genome is important, and his hope for helping cancer patients with the company’s technology.
Q: How did you become interested in the field of genetic sequencing?
Just out of grad school, I joined a small startup called Orbot Instruments that was trying to develop diagnostics not for people, but for silicon chips. I led development there, and then we got acquired by a larger U.S. company called Applied Materials. I was the general manager of their diagnostics arm. We were creating more and more advanced machines every year, with higher and higher data rates, faster solutions, more sophisticated algorithms, and more accurate robotics. I didn’t mean to stay in a big company for so long, but there was always another challenge, or another milestone to meet.
I wanted to do something new, and I wanted to see a direct benefit of what I was doing. I really looked at a lot of fields. My wife’s a doctor, my brother’s a doctor, some of my friends are doctors, and I kept seeing how data-poor medicine is. Building machines, we had sensors everywhere and diagnostics everywhere. We would know every parameter for every laser, camera, or motor. When you drive to the doctor, there is probably more information available about how your car’s engine works than there is about you. I wanted to see where ultra-high data rate diagnostics could contribute to health care.
I couldn’t find the right fit in the beginning. My “aha” moment was when I realized that in one typical blood draw, there’s about 10,000 copies of your DNA. It’s double stranded and so there’s close to 100 trillion bases. When I realized that the amount of DNA information that’s in a blood draw was roughly that number, I said, “That’s a really hard number, but it’s not infinite, it is possible.” I did something else first because I had some gaps in my knowledge. I wanted to do something else that was helpful, so I started a solar energy company that got acquired by a company called SunPower. But I kept dreaming of this. And the moment I could, I started Ultima.
Q: What makes Ultima stand out from its competitors?
If you look at the first wave of next generation sequencing companies, they mostly came from an invention or realization in molecular biology, or from a fundamental scientific understanding of something in the microbiology space. But I took a different approach. I scanned the field and tried to do a consistent, in-depth analysis of what limits the output and drives up the cost for next generation sequencing. I also tried to see what I could learn and apply from my previous industry.
When we started the company in 2016, a few years before that there was talk of third generation sequencing, moving from optical to sensing technology. I thought that sounded futuristic and promising, but I didn’t see why it was needed or how it solved a problem of cost.
Three or four others did start companies at more or less the same time I did. Very accomplished people from the biotech industry, from the sequencing industry, and they came up with different twists on today’s sequencers, usually innovations around molecular biology and chemistry. They thought that the way to take the market was to democratize it, to make accessible sequencers that you sell more of.
My belief was, if you look at the computer world, everything has gone to the cloud. If you think of biology lab testing, that’s more efficiently done at scale. So I decided to attack this market from the high end and also from the premise that this industry needs Moore’s Law [the principle that the speed and capability of computers can be expected to double every two years, as a result of increases in the number of transistors a microchip can contain] too and it should not question the demand for more and more sequencing at lower costs.
Q: What is different about your technology?
The UG 100 machine does not look like a classical sequencer, it’s a little bit larger. It uses silicon wafers, so there’s no flow cell. It’s just an open, shiny 200 mm surface, which is incredibly well engineered and clean and accurate. In traditional sequencers reagents are fed through tubes flowing through flow cells. We take a circular wafer and put the liquid we want to dispense in the center, it’s like the place where the label is on a vinyl record where there’s no music. By controlling the viscosity of the liquid and the rotations per minute, we get the thickness we want, it’s very efficient and fast. There’s also no contamination or cross docking…with flow cells there’s always a lot of washing steps, but here you don’t have any of that.
The machine looks very different, it is not designed like the current sequencers, which mostly use a kit-based approach. With kits you have to decide how big a sample you want to run and how long you’re going to run it for, how many cycles, how many bases of DNA you want to read. You then load it, click go and you come back when it’s done. With our machine, you feed it with samples, and it just keeps running. If you finish the first three, you can load three more, you can also cut in the line and make the new ones run before the old ones. It’s more of a production approach.
Q: Why is achieving the elusive $100 genome important?
The $100 genome is a proxy for the cost of sequencing. There are quite a few genomic tests happening in the world, but they’re not whole genome because it’s too expensive. The way that’s done is almost like a biological fishing process where you fish out what you want to read. Once genome sequencing is low cost enough, you just need to sequence the whole thing once.
If the doctor says, “Well, I suspect the person has this,” then the test comes back negative, then you can ask more questions of the genomic data. So that’s one trend that we see beginning to happen. Over time you will have millions of people with their whole genome sequenced to use for research if they are willing.
As I mentioned, there’s 10,000 copies of the genome in a blood draw. So even at $100 per genome, 10,000 is a million bucks. That won’t happen tomorrow, but the fact that it’s several times lower cost than it was means you can look at more copies than before.
With cancer there are DNA fragments in your body that shouldn’t be there. There’s something called minimal residual disease, which is the cancer coming back after treatment. There are trillions of cells in our body. Most are okay, but the tumor is small and it’s beginning to shed DNA. You want to catch it as early as possible in tumors that are as small as possible. By being able to read more DNA you are more likely to catch this rare event and for that you need low-cost sequencers.
Q: Who are your target users?
We are making a high throughput, short-read sequencer. Typically, you need us if you’re going to run a lot of genomes at scale, or you’re going to do a lot of liquid biopsies, or a lot of oncology testing. We think use will be mostly clinical or medical in the next few years, largely in oncology, although we’re already working with scientists working on neurodegenerative disease.
We think in the long term, every time you go for your annual checkup, they’re going to look at your cell-free DNA, because it’s information about everything, what’s happening in cells, where in your body are cells shedding, and so on. It will be like a snapshot picture of your health. We also think our machines with be attractive to people that run a lot of panels and want to move them to whole genome sequencing. There will also be some research uses like looking at ancient DNA and epigenomics.
So, why use us? First, if people want more data and second if they need higher accuracy. We are not perfect, but we are extremely good at detecting base substitutions or SNPs [F1 of 99.8%] and insertions/deletions [F1 of 99.4%]. Usually in sequencing people talk about something called a Q score to measure accuracy. For example, Q30 is 1 in 1,000 errors, Q40 is 1 in 10,000 errors. We showed recently that our machine has a Q60 score for base substitutions, which is a one in a million error rate. It’s a game changer for detecting a rare mutation.
Q: Have you had a lot of interest in the UG 100 so far?
At first, we only worked with the Broad Institute in a very secretive mode on the prototype, and later with a more mature prototype. Then we had an early access program, which was smallish, not much more than a dozen users, but really focused on the world’s leaders in sequencing. These included academic places like the Broad Institute and clinical companies like Regeneron.
There’s obviously a barrier to break when you’re trying to introduce new machines where someone’s been using something else for 15 years. There’s a lot of pain points in our customer base, for example, how many tests you can run, how good your tests are. … The fact that our machine is automated and runs in small batches all the time is just more convenient in that sense. We see a lot of motivation, but it’s going to be really interesting how it plays out.
Q: What has it been like breaking into the sequencing business and what advice would you give yourself if you could go back to when you started?
When we have an “all hands” meeting and the use cases are showing what scientists are doing with the machines, everybody’s eyes are shining with happiness that they built part of the machine or the camera and suddenly someone’s using it for cancer research. Every time I meet one of our users, I’m in awe. I think “Wow, you did that with my machine.” I came from working on advanced, complex tools and I thought they were very multidisciplinary. We had electronics, computer software, pure mechanics, robotics, imaging, and lasers, but the sequencing world is the most extreme form of multidisciplinary I’ve run into. It requires engineering, physics, microbiology, and chemistry, among other fields, so you get to work with a lot of incredibly smart folks who are from very different backgrounds, and the company does well when we all communicate well. By being able to simplify concepts and communicate across disciplines, that’s when the big inventions come up.
In physics or engineering, which is my background, you always try to think your way out of a problem. You go to the whiteboard, do resolution equations, or simulations, and maybe you’ll figure out the problem. With chemistry or microbiology, you realize you often have to experiment your way out of the problem. Investing heavily in speeding up experiments and data collection systems and making the cycle of learning results faster is often a better investment of your time than trying to magically solve the problem in your head, which tends to happen more in some other fields.
You never stop learning. And for folks who like deep tech, it’s a dream to be in this field.
Helen Albert is senior editor at Inside Precision Medicine and a freelance science journalist. Prior to going freelance, she was editor-in-chief at Labiotech, an English-language, digital publication based in Berlin focusing on the European biotech industry. Before moving to Germany, she worked at a range of different science and health-focused publications in London. She was editor of The Biochemist magazine and blog, but also worked as a senior reporter at Springer Nature’s medwireNews for a number of years, as well as freelancing for various international publications. She has written for New Scientist, Chemistry World, Biodesigned, The BMJ, Forbes, Science Business, Cosmos magazine, and GEN. Helen has academic degrees in genetics and anthropology, and also spent some time early in her career working at the Sanger Institute in Cambridge before deciding to move into journalism.
