Nov 10, 2026 8:34 AM
Heterogeneity Is the Opportunity: A Conversation with AClarity Genomics on Single-Cell Multi-Omics
Date: September 23, 2026
Participants: Jay West, PhD (Founder & CSO, AClarity Genomics), Shaun Lonergan (CEO, AClarity Genomics), Jason Amsbaugh, MBA (CEO, Samba Scientific), John Corliss, MBA (CEO, Carolina Molecular)
Overview
Single-cell genomics has spent a decade proving that cells are heterogeneous, yet it has not always been able to act on that. We recently sat down with Shaun Lonergan, CEO, and Jay West, CSO and Founder, of AClarity Genomics, which was formed to bring full-length expression and genomic, transcriptomic, and proteomic information to single-cell genomics. They explained why depth of data, not just cell count, is the next frontier, and why large language models are only as good as the data we feed them. The catch is cost: by Jay West’s math, a $100 genome sounds affordable until you need it across 10,000 cells, which makes it a million-dollar experiment.
Two Careers, One Company
Jason Amsbaugh: Jay, you worked at big organizations early on. Where did it start?
Jay West: At Sandia National Laboratories, developing biological detection technologies. Coming from small graduate and postdoc labs, I was thrust into a large organization with global goals, and I learned that technologies can have real impact, but only if they are developed well. The challenge with national labs is commercializing technology out of them. There are great stories, like QuantaLife coming out of Lawrence Livermore, but it is technology transfer, and Sandia encouraged it. That led to my first startup, which was eventually sold to Fluidigm.
Shaun Lonergan: I started in oncology laboratory testing, running in vitro chemosensitivity assays on primary tumor cells for hospitals across the US and Canada. From there I spent a career moving through molecular tools, mass spectrometry, next-generation sequencing, and diagnostics, including 454 Life Sciences, the first next-generation sequencing company, and Metabolon. When I met Jay in Raleigh, we had run side by side in the market for years without connecting. We formed AClarity Genomics to bring single-cell genomics to major opportunities in oncology and neurology, looking at full-length expression, genomic, transcriptomic, and proteomic information. We just completed a successful seed round and are working toward commercialization.
Jay West: My perfect role is Chief Science Officer. My strong suit is platform development and innovation, while Shaun focuses on the transactional and value-creation side. We have effective swim lanes. I have started a bunch of companies, and despite the ups and downs, this has been the most fun.
Matching the Data to the Question
Jason Amsbaugh: Where do you see the market need? Is this therapeutic, diagnostic, or both?
Shaun Lonergan: Top line, it plays directly into therapeutics. Full genome, full transcriptome, and proteomics bring information that drug development is going to require.
Jay West: Single cell is not important for its own sake. It is single cell to solve a problem. The market has traded the depth of data against the breadth of cells. We think every day about matching the population you need to analyze with the depth of data needed to understand the disease. Sometimes that means whole genomes, and often full-length transcriptomes, because what we really want to know is how alternative expression indicates structural changes in proteins. Take glioblastoma, which we recently studied. It is highly heterogeneous and driven by the structure of the genes, not just whether they go up or down.
When we brought the first single-cell platform to market at Fluidigm, sequencing each cell for RNA cost between $300 and $1,000. We figured out, and later published, that you could spend $10 or less, and the market expanded. But it expanded by counting whether genes go up or down, and that is not disease. Disease is dictated by structural changes, like the deletion of a kinase region that loses regulation. We have spent the better part of a decade establishing that cells are heterogeneous. The next stage is understanding what that means.
“Single cell is not important for its own sake. It is single cell to solve a problem.”
Shaun Lonergan: I started my career in chemosensitivity testing, where a drug either worked or it didn’t. This lets you understand why. Are there resistant cells in the population? What mechanism or treatment will provide the largest benefit to that patient? Side effects can be significant, so being able to treat, monitor, and re-treat on a single-cell basis matters.
Why Cost and Throughput Still Limit the Field
Jay West: These technologies cannot succeed unless they are economic. Scale and cost matter equally, and a million-dollar sequencing budget for an individual project does not exist.
Shaun Lonergan: A lot of genetic information is incomplete. End counting gives you only 30 to 50 bases of a transcript, so you see that the transcript is there but not the variants. In proteomics, quantifying mutational changes just is not there yet. And whole genome sequencing is still expensive, so many groups cannot do it. Everyone we talk to says they would love to have it. Why don’t they? Time, cost, and money.
Jay West: Throughput and cost.
Proteomics, Mass Spec, and the Single-Cell Sweet Spot
Jason Amsbaugh: Is multi-omics moving toward single-cell proteomics at scale?
Jay West: The key opinion leaders we talk to say proteins are really important. What is available today is essentially surface proteins, and essentially wild-type. Yet in our data we see modified transcripts that indicate a protein change should happen. Does it? Our advisors want to understand the functional output of all this genomics.
Shaun Lonergan: Mass spec has a real place, accurate for molecules other methods cannot handle, and it has been in toxicology for decades. But proteomics has evolved slowly because proteins are so much more complex than DNA’s four bases, with secondary structure, methylation, and sugar side chains. Doing multi-omics at the single-cell level goes hand in hand with those evolving technologies, and that is where I think the sweet spot is.
Large Language Models Need Better Data
Jay West: Large language models are fantastic at interpreting data, but they do not generate it. They are useless without data integration, and that is what we have focused on over the last six months. In some cases it has led us to what we think is a drug target in a heterogeneous tumor. We will not develop drugs at AClarity. We will enable others with better experience to do that.
When we feed a large language model data at scale, we are not getting what we are looking for, we are getting what the data is giving us. That is very different from asking a specific question. If I say I want to study five targets in a cancer, the failure rate is likely to be high, because these tumors are so dynamic, patient to patient and within a tumor.
“Large language models are fantastic at interpreting data, but they do not generate it.”
Shaun Lonergan: The metadata matters too. Getting high-quality, accurate, low-cost data into that analysis will accelerate both discovery and treatment.
Jason Amsbaugh: That is a strong case for single cell: one genome, one transcriptome, one cell, instead of detection limits set by tumor cellularity and wherever the biopsy punch landed.
Jay West: Exactly. Doctors order specific tests, but why throw away the rest of the data? We see unexpected patterns between tumor types. Pharma partners are building huge data lakes to see those intersections, and they tell us they just want more data. I do not think that is necessarily the right way to think about it, but it does make sense for feeding models. What has changed my thinking is that these cells do not change one thing. They run programs that touch how a cell migrates, differentiates, and becomes more aggressive. Large language models are exceptionally good at finding those patterns. When I first looked at them, I worried about my job. Now I think we just have to feed them better data.
Cell and Gene Therapy: Understanding the Unintended
John Corliss: Some groups use Perturb-seq to create diversity in their data. What are your thoughts on the use of Perturb-seq?
Jay West: Those methods are fantastic, but when we perturb a genome, we need to understand what we are perturbing. The promise of cell and gene therapy is amazing, but a reagent does not go to one place and make one change. Off-target effects are critically important. Perturb-seq is a great experimental model. I would be careful about modifying a genome and putting that live cell back into a human being. Understanding both the initial event and its downstream functional consequence across the whole transcript structure is where our approach could be very useful.
Shaun Lonergan: You also need to know the stability of an insert over time. Does it replicate in the same place? Do splice variants occur? Are there downstream oncological effects? It varies from tissue to tissue and patient to patient. Even HeLa cells, after five decades in labs, have shown a loss of heterozygosity and chromosomes at an astounding rate. A company’s cell line can suddenly stop performing, and you need to know why.
Jay West: How to understand the heterogeneity of these editing techniques cost-effectively has been the limiting factor for a decade, and it is what we have been thinking hard about.
Heterogeneous Tumors and Combination Therapies
Jason Amsbaugh: In diseases like glioblastoma and sarcoma, where tumors return with multiple targets, could this lead to combination therapies?
Jay West: I think so, for glioblastoma in particular. These tumors are so heterogeneous that a single target rarely tells the whole story, which is exactly why understanding the full program at the single-cell level matters.
Shaun Lonergan: We want to change “don’t go there” into “this is an opportunity.”
Jay West: St. Jude developed complex therapy programs for children that changed outcomes dramatically. Most oncology treatments are cocktails, and they were developed empirically. I think we can do it faster now.
Circulating Tumor Cells and the Economics of MRD
Jason Amsbaugh: Circulating tumor cells used to be of great interest for MRD, but cfDNA seems to have overtaken them. Are CTCs still cells of interest?
Jay West: I worked on this a lot at Fluidigm. Clinical data has been all over the place. Early-stage patients may have only a few cells, and isolating them reproducibly is hard. When these cells leave the tumor, they change their homing mechanisms, like chameleons. cfDNA is popular because it is a routine sample that, sequenced deeply, gives important information.
Shaun Lonergan: And beyond the science, if you find three cells, are you going to treat the patient? That is a cost-benefit question. Will insurers pay for monitoring? Will therapies change as a result? Can you detect them? Yes. What is the economic and social benefit? That is a market-limiting conversation.
What's Next
Jason Amsbaugh: Anything you would like to add about where AClarity is headed?
Jay West: We are super excited about what we are seeing, and more is coming really soon.
Shaun Lonergan: Jay’s team has done incredible work finding novel ways to look at very difficult problems, reducing the complexity and cratering the cost. Our partners in several key therapeutic areas are enthusiastic about the potential impact on individual patients and on society. It has not been easy, but we expect a significant reception as we roll this out in the coming years.
John Corliss: Your passion definitely comes through. You are tackling a worthwhile endeavor, and we are excited to see AClarity’s impact.
Conclusion
The thread running through this conversation is that more data is not the same as better data. Jay and Shaun argue that the next wave of single-cell work will be decided by matching depth to the biological question, by driving down the cost of that depth, and by feeding AI models data that is complete and accurate enough to reveal what disease is actually doing. Their view is that heterogeneity, long treated as an obstacle, is where the opportunity in oncology and cell and gene therapy lies.
At Carolina Molecular, we are proud to support innovators like AClarity Genomics that are helping translate multi-omic insights into standardized, actionable workflows. Whether the goal is understanding tumor heterogeneity or the unintended effects of gene editing, we share AClarity’s view that better data is what moves research toward better treatments. We invite you to follow their progress as they work toward commercialization.