Sep 21, 2026 8:59 AM
Why the Biggest Barrier in Genomics Isn’t the Science: A Conversation with Andrea Riposati of Dante Omics
Participants: Andrea Riposati (Co-Founder & CEO, Dante Omics), Jason Amsbaugh, MBA (CEO, Samba Scientific), Trent Carrier, PhD, MBA (CEO, Carolina Molecular)
Overview
Twenty years after next-generation sequencing broke onto the scene, whole genome sequencing is cheaper than ever, yet Andrea Riposati cites internal and industry data showing that fewer than 20% of oncology patients ever have their tumor sequenced, fewer than 10% of newborns are screened for genetic disease, and fewer than 1% of people with acute or rare diseases ever sequence their genome. We sat down with Andrea, Co-Founder and CEO of Dante Omics (formerly Dante Genomics), to talk about why that gap exists, how a background at Amazon shaped his approach to privacy and operations, what Dante Omics learned scaling through COVID, and where he thinks genomics, RNA, and AI are actually headed next.
From Amazon to Genomics: The Dante Omics Origin Story
Jason Amsbaugh: Looking at your background, finance and tech, what got you excited about genomics to begin with? Enough to jump in with both feet and start this company?
Andrea Riposati: Two things. The first is the opportunity to save lives and have real impact. The second is the realization that something was broken. We always talk about the potential of genomics, and everyone agrees with it, but then you look at the data: less than 20% of oncology patients use any sequencing, less than 10% of newborns are screened for genetic diseases, and even people with acute or rare diseases don’t sequence their genome, less than 1%. Genomic applications just aren’t reaching people the way they should.
There isn’t a single reason for that. You need the integration of operations, software, science, biology, and bioinformatics all together, including cost management, to offer a service that has real impact on people’s lives, with economics that let people afford it and let the company stay viable. And you need to do it globally, not just in big cities like New York and London, but in the middle of the Australian desert, or a small island in the Pacific. Today we received a genome from a small village in Uzbekistan, not even the capital. Coming from Amazon, I was excited about e-commerce, but also very aware of how important offline operations are to a successful online business. That’s really what got me excited.
The Real Barriers to Consumer Adoption
Jason Amsbaugh: What do you think have been the biggest obstacles to consumer adoption as sequencing prices kept dropping?
Andrea Riposati: Really two obstacles. The first is benefits: giving people real, actionable benefits. The second is communication: explaining to consumers why they should get sequenced, that there’s basically no risk, it’s not invasive. Meanwhile, DNA and genome sequencing are still perceived as “finding out how you’ll die,” or people ask, “why should I even do that?” Those have been the two main obstacles.
Jason Amsbaugh: Flipping that around, showing people how to live their best life feels like it has more promise, but it feels like the only people talking about that want to sell you supplements for the rest of your life. Is there a better business model for showing people how to live longer and healthier?
Andrea Riposati: There is. “Living a longer, healthier life” is a very broad statement, it can mean a better office chair, supplements, vacations. I think it’s about taking the most complex science and using it to provide real benefit. With the genome, we continuously add new reports and update existing ones based on the latest science, since what we know about the genome today is different from what we’ll know in six or twelve months. A variant of unknown significance now may be classified as pathogenic a few months from now.
What we launched this year is our RNA profile, what we’re now calling cellular monitoring: essentially whole-transcriptome monitoring to catch when your variants become activated. If you have a pathogenic variant, it will eventually become activated, and the first signal of that is RNA. Measuring the whole transcriptome every six months or so lets you catch a spike in RNA, meaning that variant has turned on, well before proteomics, and well before you’d feel sick or see a spike in a standard blood test. That’s true across healthy and less healthy populations alike. For oncology, rare disease, or common diseases like cardiovascular disease, instead of sequencing small panels with older technology, we run whole genome sequencing everywhere, customized reports based on medical history, and for oncology specifically, whole RNA plus whole exome sequencing, really using the best available technology and taking it out of a purely research setting.
A Privacy Philosophy Built on Separation, Not Sale
Jason Amsbaugh: The thing you didn’t mention that I’ve always been concerned about is privacy, the idea that a service could sell my SNP data to pharma, or that law enforcement could identify my relatives through a genealogy database. How has your approach to patient privacy been shaped by things like that?
Andrea Riposati: Our philosophy has evolved based on people’s feedback. We see the full spectrum: rare-disease patients who tell us privacy is “a privilege I don’t have,” and that they’re happy for pharma or researchers to use their data if it helps find a therapy, all the way to people who use a fake name, a fake address, and a fake email, which is also completely fine. They need the genome, they understand the benefits, and they don’t want their name attached to it.
What we’ve always done is protect the data and separate its storage: genomic data in one place, personal data in another, payment data somewhere else entirely, so it’s much harder to match those datasets back together. One of the biggest breaches of trust in genomics is selling data to an insurance company, even anonymized data, since there are ways to re-identify people from anonymized data that can create real problems, for example around life insurance. That’s a line we won’t cross. Our collaborations with pharma are fine, and people always have the choice to give research consent or not; some people even ask us to delete their data after a while, which is also fine. We essentially give everyone both GDPR-level and HIPAA-level protection, regardless of where they live. It’s better to give more privacy protection than less.
Why Cheaper Sequencing Alone Won't Drive Adoption
Jason Amsbaugh: As reporting and variant annotation have improved significantly over the last couple of years, and sequencing prices have dropped, do you feel that’s fueling growth in new customers, or is it more about awareness?
Andrea Riposati: If you look at the demand curve for sequencing, cost is a driver, but demand isn’t very elastic, you can see that in the revenues of companies like Illumina, Oxford Nanopore, PacBio, or MGI. When the cost of a genome dropped from around $600 to $200, revenue and genome volumes didn’t grow exponentially. It’s also true that the cost of sequencing is just one of many costs, if you look at some public companies’ COGS, sequencing itself is often less than 20% of total COGS, and COGS overall can be less than 25% of total costs. So a 50% reduction in sequencing cost might only be a 10% reduction in COGS.
I’d say better interpretation is actually a bigger driver, because it means you can help more people. Twenty years ago, sequencing could only help very narrow, focused patient segments. As you expand the number of diseases and clinical areas you can help with, including longevity and prevention, you expand the market, by definition, because you expand the number of people who can benefit.
Awareness is also huge, though it’s a double-edged sword, people get more excited, but some of that excitement is about supplements rather than the underlying science. There was that Moderna announcement about 10 days or so before this call: they’re sequencing the germline exome of the individual and the exome and total RNA of the solid tumor, comparing them, and using that comparison to drive a personalized vaccine. That’s a true multi-omics application, using exome and RNA rather than small panels or PCR-based tests, a great advance, both for the lives it can save and for the awareness it creates about the value of this kind of data.
Looking at my own genome, there’s pharmacogenomic information in there, I’m not currently taking those drugs, but if I ever need to, I already know where I’d need a smaller dose, or a different drug, because of a risk of adverse reaction. In my family, we had people with diabetes and kidney disease, so growing up I had all these family beliefs about avoiding cookies and chocolate. After sequencing my genome, I found out I don’t actually have a predisposition to that kidney disease or diabetes, I have other predispositions instead.
“It’s not about knowing how you’ll die, it’s about tailoring and customizing your choices for true personalized prevention.”
What COVID Taught Dante Omics About Scale
Trent Carrier: You mentioned mRNA, and I know there’s a story about Dante and COVID. Did COVID and the rise of molecular testing change public perception, and how did it change your view, or the market’s view, of what you’re doing?
Andrea Riposati: Before COVID, we’d just built our sequencing lab in Italy. We weren’t initially interested in doing COVID testing, it was chaotic, and it was hard to even find realtime PCR machines and liquid handlers. But we already had relationships with those vendors through our sequencing work, so we changed our minds and scaled up. It wasn’t easy, but it was thrilling to help, and exciting to learn to track KPIs on an hourly basis to deliver results quickly. A lot of those lessons are useful now in areas that also need scale or fast turnaround: genome-scale population programs, newborn screening, oncology.
In parallel, COVID testing was mostly real-time PCR, an older technology. Then around December 2020, when the so-called “English” and “French” variants emerged, sequencing became important, being able to predict how a virus evolves. COVID really demonstrated the value of sequencing beyond lockdowns and testing, and created more awareness of what genomics and proteomics even are. RNA is a good example: before, it was just something you studied in high school alongside DNA without really understanding the technology. Now RNA is mostly associated with mRNA specifically, and there’s also siRNA, which is something else entirely, so there’s more awareness, but also some confusion mixed in. A lot of the technology developed during that period, better lab software, better liquid handling, faster sequencing, is still in use today.
Learning from the First Wave, and What Comes Next
Trent Carrier: Genomics moved into the consumer world with 23andMe and Ancestry as the first movers. Did that inspire you, or did it feel risky? How did that first wave shape what you built at Dante, and what do you think happens with the next wave of genomics companies?
Andrea Riposati: We learned a lot from those companies, they were an inspiration, and they really innovated. Before them, there was no obvious reason why people would buy a DNA test online. It was also a real competitive challenge, since they had millions of users. We watched both companies try to launch exome-based sequencing services, which weren’t very successful, Ancestry ultimately shut theirs down, this is all public information. We respect them a lot, and I’m genuinely happy they’ve been successful. They taught people what it means to get a saliva sample at home, and kept growing even through a fair amount of unnecessary negative media coverage.
We got a lot of advice suggesting we launch a microarray service too, and decided to stick with whole genome sequencing instead. We didn’t want to be a second-rate version of 23andMe, we wanted to be the best possible Dante. We also tried to learn from Amazon’s obsession with the customer rather than obsessing over competitors. After that first wave, you now see companies growing in molecular testing more broadly, Function Health in the US, Thriva or Lucis in Europe. Distributed, at-home testing has become much more accepted by consumers. It’s easy to forget that while you can walk into a hospital or lab in New York, a lot of communities don’t have that option, so affordable, distributed solutions are genuinely valuable. We now offer some companies what we call “Genome-in-a-Box,” a ready-to-use package for launching a genome or RNA testing service under their own brand.
RNA Over Proteomics, For Now
Jason Amsbaugh: You’ve mentioned RNA a few times. What do you think about the rise of proteomic technologies, Olink, SomaLogic, Illumina, and more recently Alamar? How does that fit into a multi-omics strategy?
Andrea Riposati: We do some proteomics, but we prefer RNA and transcriptomics, both for the type of information you get and because RNA biologically comes before proteomics, at least in a clinical human setting. There are real applications for proteomics, but the economics of RNA sequencing today make it a better technology for offering a service to individuals. Epigenetics and spatial biology are in the mix too, there are a lot of technologies out there. It’s not really about the technology itself, it’s about what benefit you can create for people using it. A lot of these are still purely research-market today, and I expect a clinical, human-facing market for them within a few years, which will come down to having the right software, interpretation, and benefits.
Why Spatial Biology Isn't There Yet
Jason Amsbaugh: There’s a convergence happening between anatomical pathology and spatial biology, but it still feels cost-prohibitive, reagents, microscopy, fluorophores, multiple antibodies. What do you think drives that convergence, and does it happen soon?
Andrea Riposati: Not so soon. Even genomics took a long time to become more widespread, and it’s still not a mass-market technology today. Spatial is a good example of something still very expensive per sample, and we’ll likely need different machines and reagent economics for spatial specifically, since current platforms are optimized more for sequencing, especially germline sequencing.
I think we’ll need a single “killer application” to drive growth, the way 23andMe and Ancestry drove growth for the Illumina microarray. Years ago, Thermo Fisher was actually the preferred microarray technology, chosen by Finnish biobanks and most researchers, but Illumina became the microarray leader thanks to 23andMe and Ancestry, that genealogy application drove down microarray costs and drove a huge increase in consumption. We’ll need something similar for spatial or proteomics, maybe medical, maybe more research-oriented, maybe around food or materials, since in theory the potential for multi-omics is huge well beyond the human body: plants, food, manufacturing, new materials, viruses and bacteria in industrial applications. But today we don’t have those killer applications yet to drive costs down, and even sequencing itself isn’t a particularly efficient world: manufacturing the machines isn’t especially efficient, there’s still a lot of manual labor in labs, and they’re not typically run 24/7.
A Book You Come Back To
Jason Amsbaugh: Here’s a question I’ve wondered about as a prospective customer: I send in my genome, get it back online with browsers and multiple reports, is this going to be an interesting “book” for me to read? Or is it something I read once, and it’s mostly boring with a few key takeaways? What’s that experience actually been like for you?
Andrea Riposati: It’s probably more informative than exciting, there’s a strange dynamic where you want to find something interesting, but then you realize that finding a pathogenic variant is usually bad news. I do go back to my own genome to see updates and new reports. I don’t personally believe in applications like using your genome to choose your diet or scan a restaurant menu, but I do believe in looking at your genome when you go in for a checkup, so that information, including pharmacogenomics, is available and up to date for whoever is treating you.
Jason Amsbaugh: It sounds a bit like a user manual for a complicated piece of hardware, you don’t look at it constantly, but it’s good to have on hand for troubleshooting when you need it.
Andrea Riposati: Exactly, and hopefully you don’t need it. We’ve also had people sequence their genome for one reason and find the benefit somewhere completely different, including life-saving situations.
We had a gentleman around 60 to 65 who was concerned about Alzheimer’s and Parkinson’s. When we sequenced his genome, we found a serious oncology-related variant, and when we ran RNA, we saw a spike tied to that variant, suggesting stomach, pancreatic, or prostate cancer, it turned out to be prostate cancer. Interestingly, after his genome results, he saw his doctor at a top clinic, and standard testing there didn’t find anything. After we ran the RNA analysis, he had to go to what’s sometimes called an experimental hospital, one of those facilities in Germany where you can access technologies not yet approved by mainstream healthcare systems, and they confirmed the prostate cancer there. So this was a case where standard clinical care would have missed it, and this approach let him catch the disease at a very early stage.
There are a lot of situations like that, sometimes the benefit is immediate, sometimes it’s later, sometimes it’s peace of mind. In my own case it was less dramatic, I was worried about cookies and chocolate and sugar, and it turned out I actually needed to pay attention to iron in my blood instead, which is a fun story rather than a life-threatening one. In the end, the genome and RNA and multi-omics data remain, I’d say, a medical application. I wouldn’t change my daily lifestyle too much around it, I don’t consult my genome to decide where to go on vacation or what shoes to buy. It’s more about preventive medicine, longevity, and support for diagnosis and treatment.
Taking Genomics Directly to Community Oncology
Jason Amsbaugh: It seems like most people, at least in the US outside of academic cancer centers, still aren’t getting their tumors sequenced at all, even with first-, second-, and third-line fusion inhibitors now available. Is closing that gap in community oncology best left to companies like Foundation Medicine with large sales forces, or is that something you’d pursue more directly with doctors?
Andrea Riposati: I do believe in that direct approach. One of the core problems in genomics is that the existing technology can provide huge benefits, but almost no one is using it. Take BRCA1/BRCA2 testing: developed in the early 2000s, but in Italy and Spain it wasn’t reimbursed by the national healthcare system until 2019, and even now in Italy, to get free preventive BRCA testing as a woman, you need three cases of breast cancer in your family; two isn’t enough. That’s not because anyone involved is evil, it’s just how current healthcare systems work. We also see, privately, very wealthy, sophisticated, wellconnected individuals who’ve never sequenced their genome and don’t really know what it is, they have every means to access the best technology and simply don’t. Even at major clinics and universities, there are always a few pioneers using the latest technology, and plenty of others who are more reluctant.
We believe there’s a whole community of pioneering patients and doctors in every one of these areas, including oncology, who are off the radar of the big organizations and conference circles, and nobody’s really talking to them. While Moderna is announcing personalized cancer vaccines, plenty of people are still waiting six months for a basic PCR test or a two-gene panel. Even today, you could run pharmacogenomics via microarray to inform chemotherapy drug response, and almost none of that is being done. So we’re launching an oncology offering built specifically to work with individual doctors and their patients directly, doctor-ordered and patient-initiated, with genetic counseling included. The goal is to give people access to the best available technology without waiting for reimbursement or for healthcare systems to decide to adopt it, because if you have cancer next year, you don’t want to wait another four years for that decision.
Big Machines, Small Machines: Where Instruments Are Headed
Trent Carrier: As you look at the landscape of technology providers in genomics, and I’m not asking you to promote any particular supplier, what excites you about what’s going to drive the next wave of genomics technology?
Andrea Riposati: The first thing that excites me is anything around software and data management. All these applications, whole genome, oncology exome, somatic exome, generate far more data than a panel or microarray ever did, so anything that enables scalable analysis and reporting of that volume of data excites me, and it’s an area we invest in heavily at Dante.
On the instrument side, I think the overall market really has two applications. One is highthroughput at scale, the big machines like the NovaSeq X, running a lot of genomes efficiently on an almost fully automated workflow. The other is what we call “right here, right now,” getting results at the point of care. That’s where I’m excited about the very small machines, some of the small nanopore-based systems, as long as they’re cost-effective, including the liquid handling side, not just sequencing, and paired with fast bioinformatics and software. In theory, that could let a small clinic somewhere run a micro-sequencing lab in a small room without a big staff, just someone loading a blood sample and getting results. You’re starting to see some of that automation emerge, even from companies like Ginkgo, though a lot of it is still focused on biopharma and research rather than the clinic.
That’s what excites me, the big machines and the very small machines, for those two different use cases. A lot of the sequencing manufacturers have very broad portfolios in between that aren’t really optimized for cost, CAPEX, or turnaround time. I’d love to see more focus on the clinical space instead of just research and big universities, because I think that’s where the real benefit, and the bigger market opportunity, actually is.
Where the Real Innovation Bottlenecks Are
Jason Amsbaugh: We’re never going to get to a $500 genome if we’re still spending $200 on library prep, given how complicated PCR-based library prep and fragmentation still are. Are you seeing innovation in that upfront library-prep space, or considering engineering your own solutions in-house to solve those bottlenecks?
Andrea Riposati: We considered engineering our own solution and ultimately decided to stick with integrating existing vendors, in some cases maybe even too much. There are a lot of library-prep vendors now at different price points, and some newer technologies coming up as well. In the end, the cost of library prep today is anywhere between $10 and $100, and there are always some errors and issues that require extra reactions to account for. It’s not becoming $1 anytime soon, I don’t think anyone actually has the incentive to get it there, the same is true of flow cells. Even if sequencing itself cost $10 and library prep cost $1, your saliva collection kit would still cost around $10, shipping is another $20, and then there are facility costs, interpretation, and G&A, which can be a very large share of total cost on public companies’ income statements.
So even with a $1 genome tomorrow, you’d still need to give people a reason to buy it. Plenty of people would happily spend $1,000 on a genome today, and it already costs far less than that, but they don’t buy it because they’re not aware of it and don’t see the benefit. Even during COVID, there was surprisingly little innovation in liquid handling, library prep, or DNA extraction, despite meaningful volume. I’d love to see more applications that let non-experts run sequencing in their own labs, where someone can just put a saliva sample in a box and not think about it further, since a lot of labs today still have PhD students doing manual pipetting, which isn’t really the best use of that kind of talent.
AI's Real Role in Genomics Today
Jason Amsbaugh: Your company was early in developing automated variant reporting and genome analysis in-house. Given where the market is now, with players like SOPHiA GENETICS and Fabric Genomics, do you see continuous improvement, or is that kind of AIbased analysis starting to plateau?
Andrea Riposati: We do see improvement. We see it in interpretation, and we use AI in the lab itself, for example to balance libraries, and, going back to COVID, to help predict how a virus might evolve. We also see a real need for improvement in how doctors and everyday people actually access these technologies: today, if you hand your VCF file to an LLM directly, you don’t get useful answers back. We have a chat feature inside our genome manager, and we now let people export an annotated VCF to their own LLM if they’d like to query it themselves, since it’s already annotated, that’s much richer information than the raw file.
A lot of AI investment in healthcare right now is going into drug discovery and target identification, less into diagnostics or longevity, though there’s real potential there, and SOPHiA GENETICS is a good example of a company pushing interpretation forward with better polygenic scores and more actionable reports. But even with a perfect interpretation tomorrow, that doesn’t solve the awareness problem, and it’s sometimes focused on a very small share of the patient population, which matters enormously for those people, but the bigger gap often isn’t a technology gap at all.
“Sometimes the gap isn’t a technology gap. It’s an availability gap, a cost gap, a distribution gap, or a cultural acceptance gap.”
Conclusion
Every thread in this conversation, privacy, pricing, community oncology, AI, keeps circling back to the same point: the science has largely outrun adoption. Dante Omics’ answer hasn’t been to wait for the next scientific breakthrough to close that gap, it’s been to treat access as an engineering problem in its own right: separating data by design instead of relying on policy promises, building an RNA signal that catches disease before symptoms do, and going straight to community doctors instead of waiting for reimbursement systems to catch up.
Getting proven science to the patients who actually need it, faster than the system is built to deliver it, is a mission Carolina Molecular knows well. We’re glad to help promote the work Andrea and the Dante Omics team are doing to close that gap, and we’ll be following along as they take it further.