In our most listened to episode this year, Certis Oncology CEO Peter Ellman breaks down how his company is reinventing cancer research by building orthotopic patient-derived tumor models that more faithfully mimic human cancer — and using them to improve both drug development and treatment decisions. What is meant by orthotopic? That’s when patient tumors are placed in the “correct place” inside mice to create more faithful cancer models.

Ellman shares the deeply personal origin story behind Certis and explains why their models have changed lives. He discusses the company’s AI-driven predictive platform, now patented, that aims to double drug success rates and usher in truly personalized oncology.

Happy New Year 2026!

As sequencing continues to become cheaper, more attention is being paid to sample prep. Today we’re following up with the company, Volta Labs, a genomics applications company transforming sample prep for NGS by increasing robustness and precision, and lowering operating costs. CEO Udayan Umapathi reflects on what has been a breakout first commercial year for Callisto, the company’s sequencer-agnostic, digital-fluidics platform for sample prep. When he was last on the show, Callisto had just launched. One year later, it is deployed across North America, Europe, and Asia, with rapid uptake in clinical labs, pediatric oncology centers, and high-throughput sequencing sites.

Udayan says the scale of adoption surprised even the team. “We said we wanted to be the front end of every sequencing technology. We’ve actually done that,” he notes, adding that more than ten applications now support short- and long-read sequencing.

What’s driving the momentum? Three things keep coming up from customers: true walk-away automation, the ability to run any chemistry on any sequencer, and major improvements in quality and cost. Labs without automation engineers can now “simply buy a kit and run software…without having to learn sample prep,” Udayan explains.

A standout story this year has been pediatric oncology, where whole-genome sequencing and hybrid-capture workflows have shown strong performance on Callisto. Customers such as Prinses Máxima Center and UMC Utrecht are using the platform across Illumina, Oxford Nanopore, Ultima, and other chemistries, achieving the sequencer-agnostic vision Volta set out from the start.

Looking ahead, Udayan sees sequencing as still early in its evolution and believes sample prep has vast room for innovation. “One platform to do Illumina, one platform to do Oxford Nanopore, one platform to do Ultima… long read, short read—we do it all,” he says.

Few startups have launched with such quiet anticipation—or such a remarkable founding pedigree—as Cellanome. Backed by veterans of the genomics revolution, the company aims to do for cell biology what Illumina did for sequencing: make it measurable, dynamic, and multidimensional.

In this debut conversation, Cellanome CEO Omead Ostadan traces his path from the early days of Applied Biosystems and Solexa to what he calls “the multi-omics of the cell.” He describes a breakthrough platform capable of observing living cells in real time, combining imaging, molecular analysis, and computation in ways that bring biology closer than ever to its native state.

“Our hypothesis,” says Ostadan, “is that you are now creating an environment that most resembles the natural environment in which these cells operate. Anything you’re measuring is much more likely to resemble what you’re going to see in real biology.”

Using what the company calls CellCage technology, the Cellanome R3200 system can isolate and sustain thousands of living cells or co-cultures—neurons with microglia, for instance—allowing researchers to track interactions, responses, and phenotypic changes over time. Ostadan believes this kind of structured, longitudinal, multimodal data will be foundational for the next generation of AI-driven biological models.

“The next leap in biology,” he says, “requires a fundamentally different mode of data. That has been our focus from the start—to generate data that most closely resembles what’s happening at the foundational basis of biology across all organisms.”

Now in full commercialization, Cellanome has multiple units installed in the U.S. and preparing for expansion into Europe and Asia. For Ostadan, who has helped bring multiple life-science platforms to market, this moment feels singular:

“I’ve never been as excited about the potential of a technology as I am about what we have at Cellanome,” he says.

At the end of each year we look for a guest who in many ways defines the year. Today we sit down former NHGRI director Eric Green to reflect on the most turbulent year in his 31-year career at NIH. After leading the National Human Genome Research Institute for more than 15 years, Green’s appointment was abruptly non-renewed—a decision he learned about with “two or three days notice that I was going to have to retire from federal service.” What followed, he says, was a wave of terminations and forced retirements across NIH that left NHGRI “in trauma” as entire communications, education, and policy groups disappeared overnight.

Yet alongside this institutional upheaval, Green describes a scientific landscape moving at astonishing speed—from the maturation of genome editing and long-read sequencing to the rise of multi-omics and the accelerating push toward routine healthy newborn genome sequencing. He believes widespread newborn sequencing is no longer a distant vision but “within striking distance,” driven by global studies, new U.S. programs, and rapidly falling costs.

The conversation also explores the political pressures shaping genomics today, especially around the collection of heterogeneous genomic data and the cultivation of a diverse workforce. Green argues that scientists must learn to explain their work in human terms—as stories about patients and cures, not grants and budgets. He says it might also be a good idea to not use the “d” word (for example, “assortment” rather than “diversity”) in grants for now, silly as that is.

Despite the personal and institutional losses of the past year, Green remains committed to the future of U.S. biomedical science which continues to surge in the headlines each day. In a reference to Dickens, he says it is literally the best and worst of times.

Now entering what he calls “version 3.0,” Green sees his role as genomics evangelist, educator, and advocate—helping ensure that the momentum of genomic medicine continues even as the nation’s scientific infrastructure undergoes profound stress.

“I am officially on call to help rebuild the NIH… It’s very easy to destroy a place, and very hard to rebuild it.”

Note: This show was originally published on September 11, 2025. In light of the recent acquisition of Foresight Diagnostics by Natera, we’re re-publishing the interview with co-founders Jake Chabon and David Kurtz.

Catching a cancer relapse before any scan could see it is the ultimate goal for minimal residual disease or MRD testing. And it’s the promise behind Foresight Diagnostics, a Stanford spin-out co-founded by scientist Jake Chabon and oncologist David Kurtz who say they have arrived at “next gen” MRD testing. In this debut interview, Jake and Dave walk us through their journey from academic research to launching one of the most sensitive MRD tests on the market—one that’s already shaped new NCCN guidelines.

* 0:00 Origin story

* 4:45 What makes this “next gen?”

* 10:15 How do you get the leap in sensitivity

* 15:45 Already had an impact on NCCN guidelines

* 23:00 Launching lymphoma texting next year, then on to solid tumors

* 28:00 How will this change standard of care?

Jake explains how their novel PhasED-Seq technology, which tracks “phased variants”—usually two or three mutations on the same DNA molecule—enables unprecedented sensitivity, detecting cancer cells at levels as low as one part in 10 million. “It’s extremely unlikely to have two concurrent sequencing errors,” says Jake. “That’s functionally the core insight here.”

For Dave, who still treats lymphoma patients, the clinical need is personal. “Our goal is to treat patients until there are no more cancer cells in the body. So having a tool that tells you when there are no more cancer cells left is kind of our holy grail.”

Their MRD test, called Foresight CLARITY, launches first for lymphoma next year, with solid tumor applications in development. As their data have already begun to reshape the standard of care, Jake and Dave discuss a future in which MRD testing could come before PET scans—or even replace them.

“We want MRD testing to become the standard of care across all cancers treated with curative intent,” says Jake. With Foresight CLARITY already in three prospective trials and in NCCN guidelines, and a clear clinical need, that vision may not be far off.

This week on Mendelspod, we speak with Petter Brodin, Professor of Pediatric Immunology at the Karolinska Institutet and Director of Systems Immunology at Imperial College London, about his pioneering work in childhood immune development and his new spatial-proteomics investigations into lupus.

Petter shares how a single lecture on natural killer cells pulled him into immunology, and how early twin studies convinced him that “our immune systems are shaped predominantly by non-heritable factors.” That insight drove him to study the earliest stages of immune development—when newborns leave a sterile environment for a microbial world that imprints their immune trajectories for life.

A major theme of the conversation is Petter’s insistence that immune responses cannot be understood by looking at cells one by one. As he puts it: “Cells don’t ever work in isolation, but historically we’ve always been studying them in isolation—and I think that’s fundamentally problematic.”

This systems view is now being partly enabled by Pixelgen’s spatial interactomics. Using their Proximity Network Assay, Petter’s group is finding that lupus B cells don’t just differ in protein expression—they differ in protein distribution, revealing organization patterns that classical flow cytometry cannot capture.

These spatial signatures may point directly to new, more precise therapies. Petter explains: “If there is a difference in protein–protein interaction or protein distribution that characterizes disease, then surely that indicates a dysregulation—and that is something we can target.” Instead of broad immunosuppression or depleting whole cell populations, future treatments could focus on the exact cell states driving autoimmunity.

Petter ends on an optimistic note: spatial interactomics won’t just help treat autoimmune disease—it may allow us to intervene earlier, even preventatively, as we learn how early-life immune disturbances set the stage for disease decades later.

What if the next leap in human health isn’t hidden in our genes, but in everything that happens to them? In this week’s truly groundbreaking Mendelspod episode, we open a new chapter for the show: our first deep dive into exposomics—the study of all the physical, chemical, biological, and social exposures that shape the human body across a lifetime.

To guide us, we welcomed two leaders at the center of this emerging field: Chirag Patel of Harvard and Gary Miller of Columbia University, fresh off organizing Genomics Meets Exposomics, a landmark meeting held at the Mendel Museum in Brno—the birthplace of modern genetics. In the same abbey where Mendel tended pea plants, genomics and exposomics researchers from Europe and the U.S. gathered for the first time to build a shared roadmap for understanding how genes and environment interact to drive disease.

In our conversation, Chirag and Gary explain why the genome alone can’t answer the biggest questions in human health. While genomics accounts for roughly 20% of complex disease risk, the remaining 80% lies in our exposures—pollutants, diet, geography, stress, microbes, medications, and more—and the fingerprints these exposures leave on our biology. Exposomics, as Gary notes, is about moving from studying one factor at a time to systematically measuring the thousands of signals that accumulate in our tissues and blood.

A major theme of the discussion—and the inspiration for our episode title—is Chirag Patel’s call for exposomics to follow the same playbook that transformed genomics in the early 2000s. Just as genome-wide association studies (GWAS) revolutionized how we identify genetic contributors to disease by moving beyond one-gene-at-a-time thinking, Patel argues that the field now needs exposome-wide association studies (EWAS) to systematically search for environmental drivers. “If we are to do an exposome-wide association study… we can now discover things that were missing,” he explains, shifting from narrow, candidate-factor approaches to broad, data-driven discovery.

Both guests describe a field gaining momentum thanks to better measurement technologies, large biobanks, geospatial data, and new analytic frameworks inspired by genome-wide association studies. They also speak frankly about the remaining hurdles. As Chirag puts it, one of the major challenges is not just correlating exposures with disease but determining what these findings mean for people: “There’s a number of questions that come after that…how do you modify it? Is it causal? How do we remove it from the population if it’s adverse?”

Gary, who has spent decades studying Parkinson’s and Alzheimer’s, explains how high-resolution mass spectrometry now allows researchers to see exposure signals that were invisible before—sometimes even in decades-old blood samples. And looking ahead, he offers a clear note of optimism about exposomics’ readiness for scale: “We can do this now. It’s a reality.”

For long-time Mendelspod listeners, the episode marks an inflection point. After fifteen years covering genomics and the multi-omic revolution, this conversation shines a light on the other half of human biology—the environment—and what may become the next major frontier in disease prevention, drug development, and precision health.

After more than a decade of success in research, long-read sequencing is more and more adopted into clinical testing. In today’s show, we speak with Rita Shaknovich, Chief Medical Officer at Agilent Technologies, and Sarah Kingan, Associate Director of DNA Applications at Pacific Biosciences (PacBio), about how their collaboration is speeding up this long-anticipated transition.

* 0:00 Long read sequencing changing clinical landscape

* 7:00 Long reads replacing older technologies

* 13:15 Agilent/PacBio partnership – speeding up adoption

* 16:00 Panels designed for short reads can be used for long reads

* 24:25 Democratizing access

Long-read sequencing—once prized mainly by researchers for its ability to resolve structural variants, repeat expansions, and complex genomic regions—has reached a point of technical and economic maturity that now makes it viable in the clinical setting.

“We can now see regions of the genome that were long considered dark matter,” says Shakhnovich. “That’s leading to improved diagnostic yield and, most importantly, better outcomes for patients.”

Agilent brings to this collaboration a long-standing foothold in laboratory testing. Its automated platforms and target enrichment chemistries are already embedded in many diagnostic laboratories worldwide. PacBio, of course, brings the power of HiFi long-read sequencing to the table. Together, the companies are demonstrating that technologies originally designed for short-read sequencing can be seamlessly adapted to long-read workflows.

“Panels that were designed for short reads can be used for long reads—essentially right out of the box,” explains Kingan. “It really just opens up a whole world of clinical applications immediately.”

By combining Agilent’s infrastructure and expertise with PacBio’s long-read innovation, the partnership is accelerating the integration of comprehensive, single-platform sequencing into patient testing. The result is a streamlined, cost-effective approach that reduces the need for multiple assays while providing richer genomic insight.

What used to take months of bioinformatics analysis can now happen in minutes—and with greater biological insight than ever before. In this episode, Theral Timpson sits down with Vivek Adarsh, co-founder and CEO of Mithrl, an “AI science” company that’s bringing the power of vertical AI to the lab bench.

Adarsh began his career at Nvidia, long before the company became synonymous with AI. “What I learned there,” he recalls, “was that when you build a team around exceptional talent, deep passion, and empathy—especially empathy for your customers—everything else flows from that.” That lesson guides how Mithrl now builds tools for scientists drowning in data.

At the heart of Mithrl is a platform that takes scientists from raw data to biological insight in minutes, complete with automatic data cleaning, literature integration, and a conversational interface. Adarsh describes how one pharma team identified new biomarkers in 15 minutes—a process that would normally take months—and how another user avoided a costly error when Mithrl’s reasoning layer caught an incorrectly labeled sample.

Asked about the risk of losing “happy accidents” in a world of faster science, Adarsh pushed back:

“AI doesn’t eliminate the happy accident—it multiplies the opportunities for it. You can’t control luck, but you can create the conditions for it to appear more often.”

In closing, he offered a glimpse of what drives him:

“If we can accelerate the path from raw data to real discovery—from sequencing files to the next therapy—then we’ve done something far bigger than building software. We’ve built a partner for science itself.”

The biggest story in sequencing this year lives up to the hype. Mark Kokoris, head of SBX sequencing at Roche and inventor of the technology, joins Mendelspod to talk about how Sequencing by Expansion (SBX) works and why it may redefine the limits of genomics.

* 0:00 A long journey inspired by PCR

* 7:20 What is sequencing by expansion?

* 14:00 On scale and accuracy

* 19:40 Multi-omics vision?

* 24:40 What will be the killer app?

* 30:00 Biggest challenge for launch

Kokoris recounts the long path from co-founding Stratos Genomics in 2007 to Roche’s acquisition in 2020, when his team’s “wildly ambitious chemistry” finally found its match in Genia’s high-density nanopore platform. “Our approach to efficiently sequencing DNA,” he explains, “is to not sequence DNA. We rescale the problem—expand the molecule about 50-fold—so we can read it with much higher signal-to-noise.”

The result is astonishing speed. Working with the Broad Institute and Boston Children’s Hospital, SBX delivered whole-genome results in under four hours, with the sequencing step itself taking only about 15 minutes. Kokoris attributes the achievement to a confluence of chemistry and compute.

SBX’s duplex mode achieves Illumina-level accuracy (F1 > 99.8 %) while maintaining single-molecule simplicity. Its tunable flexibility lets small labs run a handful of samples in hours or large centers run thousands per day. Kokoris describes it as a technology built on impatience and rule-breaking, designed to give scientists options they’ve never had.

Looking ahead to the 2026 research-use launch, he’s characteristically bold:

“For me, success means SBX becoming the new standard in sequencing. Innovation can’t stop—it has to keep evolving, because biology is complex and we’ve got a lot more to do.”