Today, Verge Genomics is becoming Verge Labs, a frontier AI lab building world models of human disease biology, starting with the brain. The mission is the same, but we’re executing on it in a different way. We are going after what we believe is a fundamentally bigger opportunity, made possible by recent advances in AI. Instead of developing drugs ourselves, we are opening up our platform to other drug developers with the goal of improving clinical success rates across the entire industry.

This is not a decision we came to lightly. It is the product of a decade of building, hard-won lessons from the clinic, and a clear-eyed look at where the field is heading. Here’s how we got here.

What we built and learned from the clinic

Over the last decade, we made a bet: that an AI engine trained on human brain tissue, not mice or cells, could find better starting points for new drugs in complex diseases like ALS and Alzheimer’s disease. We called it CONVERGE. It surfaced over 280 novel drug targets, with an average 83% validation rate in preclinical follow-up. Eli Lilly and Alexion signed partnerships to use it worth $1.6 billion, and in 2024 Lilly nominated two CONVERGE-derived targets into its pipeline. The platform produced two clinical candidates of its own, including one that reached a clinical trial in ALS patients.

Along the way, we built what we believe is the largest proprietary neuroscience patient dataset of its kind in the world: more than 12,000 brain transcriptomes across 6,000 patients, 15 million single-cell profiles, matched proteomic, genomic, and clinical data, and a physical inventory of over 900 frozen brain tissue samples spanning the major neurodegenerative diseases. Last year we added one of the most thoroughly phenotyped clinical and biomarker datasets ever generated in ALS.

That first drug, VRG50635, finished its first ALS patient study in late 2025. The trial did not show clinical benefit and we are not advancing the program. We have written separately about what the trial taught us about the disease, the biology, and gaps in our first-generation platform. Read it here. The short version is that the drug hit its target in the brain and engaged the biology we predicted, but patient heterogeneity in ALS was greater than we had anticipated.

The trial taught us that finding the right target is only half the problem. The other half is knowing which patients to give it to, and how they might respond. Closing that gap is now the focus of our next chapter.

Why one drug at a time did not work

There is a version of this story where Verge keeps developing its own pipeline, takes another shot on goal, and repeats. We considered it carefully and decided against it.

The first reason is scale. The traditional biotech model was not built for a platform like CONVERGE, whose value lies in raising the probability of clinical success across a portfolio. Even a transformational platform that doubled the industry’s success rate would still see more than 80% of clinical programs fail. Proving that requires hundreds of shots on goal, not one or two.

It was also expensive. Developing a drug from target to Phase 2 takes years and tens to hundreds of millions of dollars. When capital is constrained, companies can only advance one drug at a time, and failures happen often for reasons that have little to do with the platform that surfaced the target. A platform’s value should not rise and fall on a single trial.

CONVERGE’s true value lies in the data, the platform, and the compounding learnings across many programs. The right operating model lets that learning happen at industry scale.

What’s changed since we began

Two things, in particular, have changed in the last 24 months that have informed our new direction.

The first is that the market for our kind of data has matured. Demand has picked up sharply from biotechs, from pharma teams trying to understand failed trials, and increasingly from the AI labs training domain-specific foundation models. Public deals have started to set the floor: Lilly paid Chai roughly $30M for a three-year license to a structural biology model, Noetik and GSK signed at $50M over five years for an oncology model, and Tempus AI recently booked $150M in data licensing for foundation model development to AstraZeneca. A market is forming around proprietary biological data and the foundation models built on it, and CONVERGE has more of the relevant data in neuroscience than anyone else.

The second is that recent breakthroughs in AI architectures, the same ones powering ChatGPT and self-driving cars, can now solve a problem that has long held neuroscience back: how to make use of patient data that is incomplete and fragmented.

In cancer, patients can be matched to drugs because doctors can take a sample of the tumor itself, analyze it directly, and pick the right treatment. In neuroscience the disease happens inside the brain, which cannot be sampled while the patient is alive. The field has had to rely on indirect measurements — blood, brain imaging, spinal fluid — that capture the consequences of brain disease, not the disease itself. The only direct view of brain biology comes from autopsy. And the people who donate their brains are usually not the same people enrolled in clinical trials. The two halves of the picture come from different patients.

Older machine learning approaches couldn’t bridge those halves. They required every patient to have every measurement, so the field threw out most of the available data and worked with small, complete sets. Modern generative AI can. The new generation of models learns from incomplete data at scale, combines different types of information natively, and reasons about what is missing.

But the model only works if it has brain tissue data to anchor it. Self-driving cars hit a version of this. Tesla bet that cameras alone could carry the system. Waymo and others added LiDAR — a direct depth signal that camera readings could be calibrated against — and pulled ahead. Brain tissue is the LiDAR of neuroscience, and Verge’s decade of brain tissue data is what makes the architecture work: it lets us link brain biology, longitudinal blood, and clinical trajectory into a single picture of each patient, and reason about how that patient would respond to a given drug.

What we’re building next

Our first-generation platform helped us find new drug targets. CONVERGE 2.0 goes further and includes a new wave of generative AI models — what researchers call a “world model” — that builds an internal picture of a patient’s biology and uses that picture to predict how the patient will respond to a specific drug.

We are training multimodal AI models on a decade’s worth of proprietary human tissue data together with blood, spinal fluid, and clinical data from living patients. The model learns brain biology from tissue. It learns blood biology from living patients. Then it learns to connect the two from a smaller set of patients where we have both.

The output is what we call a virtual biopsy of the brain: a reconstructed picture of what is happening inside a patient’s brain, generated from a routine blood draw. From there, the platform is designed to:

• Predict which patients are likely to respond to a given therapy

• Predict who will experience side effects before the first dose

• Predict how biomarkers will move in response to an intervention

• Identify patient subgroups inside a single diagnosis whose underlying biology is meaningfully different

• And eventually, run a patient forward in time: a virtual model of how their disease will evolve, and how different interventions would change that path

CONVERGE 2.0 is designed to catch exactly what CONVERGE 1.0 missed in our PIKfyve trial. Patient heterogeneity, predicted from a baseline blood draw before any patient is dosed. Brain state, inferred from the peripheral biomarkers we can actually collect from living patients. The direction a biomarker will move under a given intervention, predicted in advance rather than discovered after the trial reads out.

Together, these capabilities let us predict not just which targets to pursue, but which patients to pursue them in.

The business

Underneath all of this is a bet about where AI value will compound over the next decade. As general-purpose AI models hit the limits of the public text for new training data, the next leg of progress will increasingly come from vertical AI: domain-specific models trained on proprietary data, with the expertise to make sense of it. Life sciences is among the largest of those verticals.

High-quality multi-modal patient data will become as essential to drug development over the next decade as sequencing data was over the last one. AI labs need it to train their models. Drug developers need it to design better trials. Regulators are increasingly asking for it. The companies that supply that layer will sit at the center of how the field works, and that is the layer Verge has spent a decade building.

Drug developers can engage with CONVERGE today in whichever form fits their question. Some bring a specific translational problem (how their target behaves in human tissues, which patients to enroll, how to stratify a subtype) and we run CONVERGE against it. Some license drug targets we have already discovered in indications where they have downstream capabilities we do not need to build ourselves. Others license our data and AI models directly, or build new models with us. These are three doors into the same datasets, and every engagement adds to the underlying dataset. Over time, the data and AI layer itself becomes the primary product.

The vision

In the near term, we are working toward precision neuroscience: matching patients to drugs based on what is happening inside their brain, rather than what is visible at the surface. Smaller trials. Higher success rates. Drugs that work because they are given to the right people. This is the shift oncology made a generation ago, and it is overdue for the brain.

The longer-term vision is more ambitious. Once we integrate longitudinal data, our foundation models will be able to predict patient trajectories. A virtual model of each patient, run forward in time, to see where the disease is heading and how different interventions would change that path.

Transformational platforms are not built overnight. They come from persisting after setbacks and iterating on hard-won lessons. We come into this moment with a decade’s worth of tissue data acquisition and neuroscience drug development experience, a proprietary human data atlas across more than 6,000 patients, and the experience of taking an AI-discovered drug all the way through a clinical trial in patients. Few AI-native biotechs have lived that arc. It is the foundation Verge Labs is built on.

The work itself has not changed. Read human biology directly, in patients and in disease tissue, and use what you learn to design better drugs. What has changed is the scope. Doing this for one drug at a time means we will only ever learn from what we built ourselves. Doing it across a field means hundreds of opportunities to build knowledge around what works. That scale is what Verge Labs is built for.

When a drug doesn’t work, the temptation is to look away. We’re going to do the opposite. The trial taught us more than a successful one might have, and we want to share what we learned.

Our first generation discovery engine, CONVERGE 1.0, nominated a drug target called PIKfyve. Our team took it through years of preclinical work, Phase 1 healthy-volunteer studies, and a Phase 1b study in ALS patients. The drug had clear biological activity, including in the brain and the two cell types most central to ALS, neurons and astrocytes. But the patients did not get better. Inside the gap between those two facts, we learned things about the disease and our platform that change how we build the next iteration.

We got very close. We were in the right neighborhood. We just have not yet found the right house yet. Here is what we learned, and what we are going to do with it.

What we set out to do

PIKfyve is an enzyme that helps cells take out the trash. It is part of the endolysosomal system, the machinery cells use to move, recycle, and dispose of internal cargo. From a multi-omic analysis of nearly 1,000 tissue samples from ALS patients and healthy controls, CONVERGE flagged PIKfyve as a top target. In ALS brain tissue, the gene network around PIKfyve was profoundly dysregulated. Motor neurons in ALS could no longer clear the protein aggregates piling up inside them, leading to cell death. The house was full of trash. Restoring the pathway should help take it out and keep the neurons alive longer.

In motor neurons grown from ALS patient stem cells, PIKfyve inhibition cleared protein aggregates and improved survival. In a mouse model of ALS, it reduced neurofilament, the same blood marker we’d later use to measure efficacy in the clinic. Independent labs saw similar effects across multiple preclinical models (Shi et al., Nature Medicine, 2015; Hu et al., Cell, 2023). The biology was novel, the genetics were supportive, and the case for translation was strong.

We developed VRG50635, a potent, brain-penetrant PIKfyve inhibitor, and ran Phase 1 healthy-volunteer studies. We concluded that the drug was safe and well-tolerated. We identified plasma GPNMB as a reliable pharmacodynamic biomarker: a readout confirming the drug was hitting the PIKfyve pathway. We then ran a Phase 1b in 54 patients with ALS powered to detect efficacy, which we defined as a 30% drop in a protein called plasma neurofilament light chain (NfL).

NfL is a protein neurons release as they die. A fall in NfL levels in the blood is widely read as neurons dying more slowly, which should translate into slower disease progression. It is a widely used biomarker of efficacy in ALS clinical trials.

Pharmacologically, the drug behaved as designed. It reached the brain at concentrations roughly twice what’s needed to fully inhibit PIKfyve. GPNMB in blood and cerebrospinal fluid (CSF) rose in proportion to drug exposure: precisely the response one would expect if the drug were hitting its target.

Then the data showed something we did not anticipate.

The two surprises

Plasma NfL went up, not down

We had assumed plasma NfL would fall once cellular trash clearance turned on. Instead, plasma NfL rose. The signal was detectable within 2 weeks of starting the drug and stayed elevated as long as patients were on it. Plasma GFAP, another protein from inside the same cells, climbed six hours after the first dose and tracked with NfL throughout the trial.

Spinal fluid told a different story. CSF NfL trended down, with a larger drop in patients with the C9orf72 genetic form of ALS. A reduction in NfL in CSF is considered early evidence of clinical benefit. But the primary efficacy endpoint was a reduction in plasma NfL, and plasma did the opposite of what we had bet on.

We believe the rise in plasma NfL was a pharmacodynamic response to PIKfyve inhibition — a direct effect of the drug’s mechanism, not worsening disease. Patients on drug did not get worse faster, and when dosing was interrupted, plasma NfL fell rapidly, on a timescale that matches drug clearance, not recovery from neuronal injury.

The biology makes sense in hindsight. PIKfyve inhibition activates a cellular pathway (secretory exocytosis) that actively pushes intracellular contents out of the cell. The drug was opening the door to expel the trash inside. Misfolded proteins from neurons from inside diseased cells came pouring into the bloodstream, including NfL from neurons and GFAP from astrocytes. Plasma NfL was tracking the drug’s action, not the disease. Unfortunately, it was also our efficacy endpoint.

NfL, one of the field’s most relied-upon efficacy biomarkers, behaves in a mechanism-dependent way when an intervention engages the cell’s recycling system — a finding we did not anticipate. Given that the endolysosomal pathway is of high-interest in many neurodegenerative diseases, it’s one we think the field should be aware of.

Without a reliable efficacy biomarker, we had to rely only on clinical endpoints, where no effect was seen within the constraints of the trial.

ALS patients behaved differently

Our first Phase 1 healthy volunteer study was uneventful. In healthy people, the drug was well tolerated at all doses, except at very high exposures.

Then came the Phase 1b trial in ALS patients. To our surprise, a third of ALS patients could not tolerate even the lowest dose and discontinued before any dose escalation. But the other two-thirds escalated through higher doses and looked much like the healthy volunteers. Why?

The split tracked with disease severity. Early discontinuers tended to be older, with more advanced ALS and worse prognostic scores. They responded differently to the same drug despite identical pharmacology. Something in their underlying biology was driving the split.

We measured 48 baseline lab values and biomarkers. Two — and only two — were significantly different between the two groups: plasma NfL and plasma GFAP. Both were elevated at baseline in the patients who did not tolerate the drug. The same two proteins our drug was pushing into circulation.

Severe ALS patients here seem to be doing the PIKfyve thing on their own. Their cellular recycling system is already running flat out, presumably as a response to the misfolded protein piling up in their neurons. We came in with a drug that pushed the same dial the disease was already pushing. For the sickest patients, there was no dial left to turn, only side effects.

A similar pattern shows up in immunology. IL-10 is one of the most potent anti-inflammatory cytokines in the body. It has never been approved as a therapy, partly because acutely sick patients are already producing high levels of IL-10 on their own. The pathway is saturated. The drug has no room to work.

PIKfyve dysregulation is not a yes-or-no feature of ALS; it tracks with how sick the patient is. Some patients are at the start of that arc, some are near the end, and they likely need very different drugs, very different doses, or possibly the opposite intervention entirely. We treated them as one population. The trial taught us they are not.

What the platform got right

Even though the drug did not deliver clinical benefit, it’s worth examining where we were able to move the needle.

The drug moved real biology in the cells the disease actually lives in. Most ALS trials end with no biological signal at all. In ours, the drug crossed the blood-brain barrier, engaged the pathway, and triggered rapid release of NfL and GFAP. Since NfL comes from neurons and GFAP from astrocytes, the drug was clearly triggering biological activity within the two cell types most central to ALS.

We could prove the drug reached its target, something many past ALS trials lacked. In our PIKfyve network from human tissue, GPNMB stood out as one of the most tightly coupled genes to PIKfyve. In ALS patient blood and spinal fluid, we found GPNMB was released by cells, detectable in both, and elevated in disease. The drug raised GPNMB across cells, ALS patient-derived motor neurons, and rodents, in lockstep with how much drug was on board. In healthy volunteers, the same pattern held in blood and spinal fluid (Gontier et al., Clinical and Translational Science, 2026). That gave us a clear readout that the drug was hitting its target in both the body and the brain before we dosed a single ALS patient, and a way to set the right dose for the Phase 1b.

The platform predicted new ALS-specific biology. In the clinic, the PIKfyve pathway turned out to be profoundly different not only between ALS patients and healthy volunteers, but within ALS itself. The sicker the patient, the more their cells had already activated the PIKfyve pathway in response to the disease, and the more strongly they reacted to the drug. This points to a new, disease-relevant biology: in ALS, cells appear to turn up the PIKfyve pathway as a compensatory defense against the disease, with the response scaling alongside severity.

Finding biology that tracks with a disease is one thing. Knowing which patients to treat, and when, is another. That is the gap CONVERGE 1.0 left open, and the one CONVERGE 2.0 is built to close.

What’s next and why now?

ALS is one of the hardest diseases in drug development. Even if a platform doubled the industry’s clinical success rate (a transformational result), more than 80% of programs would still fail. We are not guaranteeing success on any single asset. We are raising the probability of success across a portfolio.

Transformational platforms are not built overnight. They come from persisting after setbacks and iterating on hard-won lessons. We have done what many companies in our space have not: taken a novel target from an AI platform through to a clinical proof of concept trial with clear biological activity. Along the way, we accumulated what we believe is the largest proprietary, curated neuro tissue dataset in the world — more than 12,000 human tissue samples across 6,000 patients and 15 million single-cell profiles, paired with clinical and genetic annotations across ALS, FTD, PSP, Parkinson’s, schizophrenia, and others. The PIKfyve trial now adds one of the most thoroughly phenotyped clinical and biomarker datasets ever generated in ALS.

What has changed in the past two years is the AI itself: recent breakthroughs in architecture have arrived just as our dataset has reached the scale they need. The hardest problem in neuroscience data is its incompleteness. Brain tissue is the molecular ground truth for neurodegenerative disease, but it can only come from autopsy patients. From living patients we get accessible proxies for the brain: blood, imaging, and clinical data. The two halves of the picture come from different patients.

Modern generative AI can bridge that gap. It learns relationships across modalities from many patients and infers what is missing. But the model only works if it has brain tissue to anchor those relationships. Without it, the proxies show only the disease’s downstream effects, with no way to trace them back to what is actually driving the disease in the brain. Our decade of brain tissue data is what makes the architecture work: it lets us link brain biology, longitudinal blood, and clinical trajectory into a single picture of each patient, and reason about how that patient would respond to a given drug. That is the class of questions the PIKfyve trial asked and we could not answer.

CONVERGE 1.0 was a discovery tool, built to predict novel targets from human data in the toughest diseases. We are now expanding CONVERGE 2.0 to include a precision neuro model, by training the latest generative AI architectures on a decade of multimodal neuroscience patient and clinical data. The output is what we call a virtual biopsy of the brain — an inferred molecular picture of a patient’s brain, generated from a routine blood draw. From there, it is designed to match patients to the right therapies, predict how patients will evolve over the course of their disease, and predict how biomarkers will respond to different interventions.

In other words, CONVERGE 2.0 is designed to catch exactly what CONVERGE 1.0 missed. Patient heterogeneity, identified from a baseline blood draw before any patient is dosed. Brain state, inferred from the peripheral biomarkers we can actually collect from living patients. And the direction a biomarker will move in response to a given intervention, predicted in advance rather than discovered after the trial reads out.

Together, these capabilities let us predict not only which targets to drug, but which patients to drug them in.

We were in the right neighborhood. Fortunately, the map is about to get considerably better.