ALS Gene Therapy: How a Genomic Discovery Becomes a Drug
Written By: Ryan Morrison
Listen On: YouTube | Spotify | Apple Podcasts
2026 Asset Allocation Report
See how 230+ HNW investors with an average net worth of $17M are allocating across public equities, private markets and alternative assets.
For most people diagnosed with ALS, the treatment options available today are not meaningfully different from what Lou Gehrig faced in 1939. The ALS Association reports that average survival after diagnosis is roughly three years, and no drug has yet changed that trajectory for the majority of patients. That is the starting point for Trace Neuroscience, a clinical-stage biotech company targeting a protein called UNC13A — first identified in uncoordinated worms in the 1960s, now the basis for a genomic medicine designed to restore what ALS patients lose.
Eric Green is the CEO of Trace Neuroscience, a cardiologist by training with an MD-PhD and over a decade building genomic medicine companies alongside venture capital firms. In a conversation on Navigating Wealth, he walked through how a basic science discovery becomes a drug, why biotech companies create value long before they generate revenue, and why the AI-enabled drug discovery revolution is still further away than most people assume. Before getting into his work, the hosts also discussed a question that generates more peer debate inside Long Angle than almost any other: whether to leave significant wealth to your children, and how to think about the structure when you do.
TL;DR
Long Angle members discussed whether to leave significant wealth to children, with most leaning toward supporting specific life moments over lump transfers — and Charlie Munger's counterpoint getting the last word.
ALS has no meaningful treatment for most patients despite decades of research. Genomic sequencing changed that by identifying proteins like UNC13A that drive disease progression.
UNC13A was first found in uncoordinated worms in the 1960s. Trace Neuroscience is now using a genomic medicine — a nucleic acid sequence — to restore the missing protein in ALS patients.
Biotech companies create value through clinical risk reduction, not revenue. A single trial readout can move a company's value more than years of operations.
AI is useful for document prep and antibody optimization in biotech, but has not yet changed target identification or clinical development — the parts that matter most.
The COVID vaccine's speed came from a platform built over years, not from regulatory leniency. That lesson applies directly to how ALS trials are now being designed.
Why ALS Still Has No Effective Treatment — and What Changed
ALS has resisted meaningful treatment for over a century because most cases do not trace back to a single genetic cause, which made traditional drug targeting approaches applicable to only a small minority of patients.
The disease attacks motor neurons in the brain and spinal cord, progressively eliminating the ability to move, speak, swallow, and breathe. Until recently, the two approved treatments — riluzole and edaravone — modestly slowed progression in some patients but did not address the underlying mechanism. For most people with ALS, the disease course remained essentially unchanged from what it was when Lou Gehrig retired from baseball in 1939.
What changed was the sequencing of the human genome. Once researchers could compare the genomes of large populations — people with ALS against people without — they could identify which genes confer added risk, which variants are associated with more severe disease, and which proteins might be tractable targets for intervention. This is population genomics applied to drug discovery, and it is what led Eric Green and his collaborators to UNC13A.
What UNC13A Is and Why It Matters for ALS
UNC13A is a protein essential for communication between nerve cells and muscle cells. When it is present and functional, neurotransmitters are properly released at synaptic junctions. When it is depleted, that communication breaks down. A 2022 paper published in Nature from teams at UCL and the NIH established that TDP-43, a protein whose dysfunction is present in roughly 97% of ALS cases, drives the depletion of UNC13A in affected neurons. Genomic risk variants in UNC13A compound this effect — people who carry those variants lose UNC13A faster and experience more severe disease.
The implication is significant: UNC13A loss is not a rare familial ALS phenomenon. It is linked to the same TDP-43 dysfunction found in the overwhelming majority of ALS cases. That makes it a different kind of target than what most prior ALS gene therapies have pursued.
Beyond Wealth Newsletter
How HNW founders and executives navigate the questions wealth creates — grounded in peer data and Long Angle community discussions. Free, delivered every Thursday.
How a Protein Found in Worms Became a Clinical-Stage Drug
The path from the UNC13A discovery to a clinical trial spans six decades and two entirely separate research traditions — and it illustrates how the most commercially significant medicines almost never start with commercial intent.
In the 1960s, researchers were systematically mutating every gene in the roundworm C. elegans to see what happened. When they mutated the gene that would become known as UNC13A, the worms became uncoordinated — they could not move properly. Over the following decades, researchers established why: the UNC13A protein is essential to how nerve terminals communicate with muscle cells, enabling the docking and release of neurotransmitters at synaptic junctions. Loss of that protein means the signal never arrives.
That basic insight sat in the scientific literature, relevant to worm biology and eventually to broader neuroscience, until the genomic era produced the population-scale data needed to connect it to ALS. Green and his collaborators were involved in the research that identified UNC13A as a driver of ALS severity in humans — confirming in post-mortem brain samples and lab-grown neurons that the same protein loss seen in uncoordinated worms was present in ALS patients.
From that discovery, Trace Neuroscience's approach is direct: design a medicine that restores the missing protein. The mechanism is a genetic medicine — a nucleic acid sequence engineered to bind to the RNA that makes UNC13A and ensure healthy levels of the protein are produced in affected neurons. This is not a gene silencing approach. It is a restoration approach, targeted at giving back what the disease takes away.
Why Genomic Medicines Can Move Faster Than Traditional Drug Development
The COVID vaccine provided the most public demonstration of this dynamic. Once the sequence of the SARS-CoV-2 virus was known, designing an mRNA vaccine took days — not because regulation was relaxed, but because the mRNA platform had been built over years. The design step, which historically consumed enormous time and resources, became nearly instantaneous.
Genetic medicines for diseases like ALS work similarly. Knowing the target sequence — the specific RNA for UNC13A — allows the medicine to be designed with high precision. The years of work come in manufacturing the molecule to pharmaceutical standards, validating it in animal models, and building the regulatory package. But the target specification step, which in traditional drug development involves enormous uncertainty, is largely resolved by the genomic discovery.
How Biotech Creates Value Before It Has Revenue
Biotech companies are valued not on sales but on how much scientific risk they have removed from a thesis — which is why a single clinical trial readout can move a company's value more than years of operational execution.
As Green described it, one of the most disorienting aspects of biotech for people coming from other industries is that none of the companies he has built have ever had to deal with the burden of revenue. That is not a failure state — it is the standard model. Companies in this space are regularly acquired or partnered by large pharmaceutical firms at valuations that reflect the probability-weighted value of drugs that are still five or ten years from reaching patients.
What drives that value is derisking. Each successful experiment — each animal study that confirms a drug behaves as intended, each clinical readout that shows safety and early efficacy signals — reduces the probability that the scientific thesis is wrong. Value in biotech is created by reducing technical risk, not by generating sales, which is why the binary character of clinical readouts can produce such dramatic valuation swings.
Platform Biotech vs. Focused Development Companies: What the Distinction Means for Investors
Green described a spectrum of biotech company types that is useful for anyone trying to evaluate this as an asset class. At one end are platform companies — built around a broad technology (CRISPR is the clearest example) with the thesis that multiple medicines can emerge from the same underlying capability. These companies raise large amounts of capital early, aggregate intellectual property and scientific talent, and operate on long timelines. At the other end are focused development companies, which exist to answer a single question: can this specific molecule be developed into a meaningful medicine, either in a new indication or through a different development approach?
The current market has shifted meaningfully toward the focused end of that spectrum. Investors have become more skeptical of early platform risk — the uncertainty about which molecule to pursue and whether the platform can actually generate drugs — and more interested in companies that have already defined the scientific question and are raising capital specifically to answer it. This has implications for how biotech is financed and how long it takes companies to reach the binary events that drive liquidity.
Navigating decisions where peer context matters most?
Long Angle members include founders, operators, and investors who compare notes on private markets, estate planning, and the decisions that most financial resources aren't built for.
The community also happens to include people like Eric Green — practitioners building inside the industries where capital gets deployed.
The Academic-to-Biotech Pipeline and Why It Is Under Pressure
The discoveries that become transformative medicines almost always originate in curiosity-driven academic research with no commercial intent, and the funding structure that supports that research is changing in ways that have direct consequences for what the biotech pipeline looks like in a decade.
Green's framing of the ecosystem is worth internalizing: basic research provides the discoveries. Biotech provides the resources — financial, scientific, and operational — to turn those discoveries into drugs. When both work, they are genuinely complementary. When the basic research foundation weakens, the pipeline that feeds biotech weakens with it.
CRISPR is the model case. The gene editing technology that now underpins some of the most commercially significant medicines in development originated in research into how bacteria defend themselves against viruses — research funded by government grants with no pharmaceutical application in mind. Doudna and Charpentier's Nobel Prize-winning work emerged directly from studying bacterial immune systems. No one in biotech or pharma was funding that research. The academic infrastructure was.
What the CRISPR Discovery Teaches Us About Where the Next Platform Comes From
The current pressure on academic research funding is not primarily about the grants that go directly to individual investigators and their labs. It is about the overhead funding — the money that goes back to research institutions to support the physical infrastructure, administration, and operational capacity that makes those labs possible. Green was direct that the structure of those agreements has no first-principles justification; it evolved that way, and changing it will be painful for institutions that have built their cost structures around it.
What this means for investors evaluating biotech funds and platform companies over a ten-year horizon is worth considering: the basic science discoveries that become the next generation of medicines are being made, or not made, right now in academic labs. The UNC13A story required decades of worm biology, genomic sequencing infrastructure, and clinical data collection — none of which a biotech company was going to fund on its own.
Is AI Actually Changing Drug Discovery?
AI is improving efficiency in biotech's back-office and some specific technical applications, but it has not yet meaningfully changed the target identification and clinical validation process where the highest-value decisions happen.
Green's assessment was direct about this. The day-to-day work of running a biotech company — regulatory document preparation, literature review, operational tasks — does benefit from the same AI productivity gains seen elsewhere in knowledge work. But even these near-obvious applications face friction in a heavily regulated industry: introducing AI into regulatory document preparation makes compliance teams nervous in ways that slow adoption even when the logic for it is sound.
The bigger claim — that AI will transform drug discovery itself by identifying new targets or designing better molecules from scratch — remains largely unrealized. Green identified the one area where early progress is genuine: once a drug target is known, AI can help optimize the specific molecule designed to hit it. In antibody development, for example, AI tools are beginning to accelerate the process of identifying which molecular variants are most likely to be effective. That is meaningful, but it is a narrow slice of the full drug development pipeline.
The Feedback Loop Problem: Why Drug Discovery Doesn't Work Like Software Development
In software, an AI system can iterate rapidly because feedback is immediate. You can test a hypothesis, see the result, and adjust within hours or days. Drug discovery does not work this way. Identifying a target and then building, testing, and validating a molecule that hits that target in humans takes years — sometimes decades. The feedback loop that would allow AI to learn from its predictions is stretched across timescales that make rapid iteration impossible.
Green noted that one of the most appealing theoretical applications of new genetic technologies is shortening this gap — reducing the time between identifying a target and being able to run the human experiment that tests whether the hypothesis is correct. That compression is more likely to come from the genetic platform infrastructure than from AI, at least in the near term.
How the COVID Vaccine Changed What's Possible — and What It Didn't
The COVID vaccine's seven-to-ten-month development timeline was not primarily a story about regulatory flexibility — it was a story about what happens when a programmable platform technology meets a defined target.
The sequence of the SARS-CoV-2 virus was known almost immediately. Because the mRNA platform had been built over years before the pandemic, designing a vaccine that incorporated that sequence was a matter of days, not years. The manufacturing infrastructure had also been developed in advance. What the emergency compressed was validation and approval timelines — real compression, and real risk-benefit recalibration — but the scientific foundation had been laid long before COVID-19 existed.
The lesson Green drew from this maps directly to the ALS work: the path to faster development runs through better platforms and better basic science, not primarily through lighter regulation.
How ALS Trial Design Reflects a Different Risk-Benefit Calculation
The FDA's approach to severe, fatal diseases like ALS reflects a proportionality logic that differs from how most drug approvals work. Standard drug development typically begins with healthy volunteers at very low doses, designed primarily to confirm safety rather than demonstrate efficacy. For ALS patients, that approach carries a cost that is hard to justify.
Green's framing of this was precise: for most people with ALS, they only get one clinical trial. The typical life expectancy is two to three years from diagnosis. If a patient enrolls in a trial at a dose too low to plausibly have any effect, they have used their one opportunity without any reasonable chance of benefit. Trace's clinical approach is to start patients at doses that could plausibly work and to treat them for long enough that clinical benefit — if it is present — can actually be detected. The FDA's willingness to support that design for severe life-threatening diseases is a meaningful shift in how the agency applies its risk-benefit framework.
The first patient treatment in Trace Neuroscience's human clinical trials is planned for 2026. Whether restoring UNC13A protein levels translates into clinical benefit for ALS patients is the question that trial is designed to answer.
Frequently Asked Questions
What is ALS gene therapy and how does it differ from traditional ALS treatments?
ALS gene therapy targets the genetic or molecular mechanisms driving motor neuron loss, rather than managing symptoms after they appear. Most current ALS treatments — riluzole and edaravone — modestly slow disease progression in some patients but do not address the underlying cause. Trace Neuroscience's approach targets UNC13A, a specific protein whose loss is now linked to the TDP-43 dysfunction present in most ALS cases, using a genomic medicine designed to restore healthy protein levels in affected neurons.
What is antisense oligonucleotide therapy for ALS?
An antisense oligonucleotide (ASO) is a synthetic nucleic acid engineered to bind to a specific RNA target and alter how that target behaves — either suppressing a harmful protein or, as in Trace's approach, restoring a missing one. ASOs have already shown clinical benefit in related motor neuron diseases including spinal muscular atrophy, where an ASO-based drug called nusinersen has substantially improved outcomes. For ALS, ASO approaches targeting SOD1 (tofersen, the first FDA-approved ALS gene therapy) and other genetic drivers are in various stages of development.
Why do most ALS patients have so few treatment options?
Most ALS cases are sporadic — meaning they occur without a clear family history or identifiable single-gene mutation, making standard genetic targeting approaches applicable to only a small fraction of patients. The UNC13A target is different because it is connected to TDP-43 dysfunction, which is present in roughly 97% of ALS cases, not just familial forms. That broader applicability is part of what makes the Trace Neuroscience thesis scientifically significant.
How does a biotech company get acquired before it has revenue?
In biotech, value is created by reducing scientific risk, not by generating sales. A successful clinical trial readout — even a small Phase 1 or 2 study showing that a drug is safe and produces the expected biological effects — can make a company highly valuable to a large pharmaceutical firm that has the commercial infrastructure to take the drug the rest of the way. These acquisitions and partnerships routinely happen five to ten years before any drug reaches patients, and they represent the primary liquidity mechanism for biotech investors.
What role does academic research play in biotech drug development?
Virtually every transformative medicine in the last fifty years traces back to basic academic research that had no commercial purpose at the time it was conducted. CRISPR originated in the study of bacterial immune systems. GLP-1 obesity drugs originated in research on the Gila monster's digestive physiology. The mRNA vaccine platform was built on decades of academic RNA biology. The academic-to-biotech handoff is not incidental — it is the primary source of genuinely novel drug targets. When basic research funding weakens, the pipeline that feeds biotech weakens with it, on a delay that may not be visible for years.
How is the FDA approaching rare and fatal disease drug development differently?
For severe, life-threatening diseases with no effective treatments, the FDA is applying a proportionality framework that reflects a different risk-benefit calculation. Clinical trials for diseases like ALS can begin in patients rather than healthy volunteers, at doses designed to plausibly produce benefit rather than purely establish safety at sub-therapeutic levels. The agency has also indicated willingness to support development of N-of-1 therapies — medicines designed for individual patients with ultra-rare genetic diseases — under a more flexible regulatory pathway than would apply to drugs intended for broad populations.
Is AI changing how new drugs are discovered?
AI is improving efficiency in specific parts of the drug development process — particularly in antibody optimization once a target is known, and in back-office tasks like document preparation. But AI has not yet meaningfully changed target identification or clinical development, which are the highest-value and highest-risk parts of the pipeline. The core constraint is the feedback loop: testing a drug hypothesis in humans takes years, which prevents the rapid iteration that makes AI powerful in domains like software. Green's view is that genetic platform technologies are more likely to compress that timeline than AI, at least in the near term.
Final Thoughts
The path from a protein found in uncoordinated worms in the 1960s to a clinical trial in ALS patients in 2026 is not a story about AI or regulatory shortcuts. It is a story about what happens when curiosity-driven basic science, focused company-building, and a disease-first clinical philosophy operate over long timescales. The genomic tools that made the UNC13A discovery possible were built on decades of academic research that was funded without any particular drug in mind. The platform technologies that will make genetic medicines programmable — the lesson the COVID vaccine taught us — were also built that way.
For the Long Angle audience, Eric Green's conversation is useful on two levels. The specific biotech development model — how companies are structured, how value is created before revenue, what a binary event actually is and why it matters — is decision-relevant context for anyone evaluating biotech exposure in an alternatives portfolio. The broader point about where medicines come from, and what threatens that, is relevant to anyone who wants to understand what the next decade of drug development will actually look like.
The kind of peer intelligence that put this conversation on the table — a practicing biotech CEO, a community member, discussing both wealth transfer and genomic medicine in the same hour — is what Long Angle is built around.
Members include people who have done the thing you are thinking about, in domains that span investments, exits, estate planning, and the industries where capital actually goes. Long Angle is a vetted community of high-net-worth founders, executives, and investors. Members access peer-led investment diligence, candid discussions on complex wealth decisions, and the kind of first-hand context that doesn't come from advisors or financial media.
Resources Mentioned
Trace Neuroscience — Eric Green's company, developing a genomic medicine for ALS targeting UNC13A
ALS Association — Understanding ALS — patient-facing overview of disease progression and survival data
UCL News — UNC13A Discovery (Nature, 2022) — the peer-reviewed research establishing UNC13A's role in ALS
