Model Medicines Discovers the RdRp Thumb-1 Pocket as a Conserved Target for Broad-Spectrum Antiviral Development

Discovery described as a “fundamental shift in antiviral drug development” by an expert peer reviewer; serves as the basis for an armamentarium of broad-spectrum, direct-acting antivirals.

SAN DIEGO — Model Medicines, an AI-first biotechnology company developing first-in-class therapeutics against multi-indication biological choke points, today announced the publication of a manuscript that establishes the RNA-dependent RNA polymerase (RdRp) Thumb-1 pocket as a conserved, druggable target for broad-spectrum antiviral development.

The Thumb-1 pocket is a cryptic, allosteric site on the viral RdRp that controls an essential conformational change required for polymerase initiation. Pharmaceutical engagement traps the viral polymerase in a catalytically incompetent state, inhibiting replication. 

“The scientific literature states that conserved, druggable, allosteric viral biology targets do not exist,” said Daniel Haders, PhD, Founder and CEO of Model Medicines. “The broad-spectrum RdRp Thumb-1 pocket disproves decades of dogma that has influenced and continues to influence every antiviral drug program. We discovered a target the field dismissed as impossible. We did this by methodically developing a platform that allows us to do the tedious work of target discovery that others will not do or cannot do.”

The preprint, “The RdRp Thumb-1 Pocket is a Conserved Target for Broad-Spectrum Antiviral Development,” is available on bioRxiv. It is accompanied by a companion preprint reporting preclinical proof of concept for MDL-001, the company’s oral, direct-acting antiviral discovered against this target.

Model Medicines has been invited to present the RdRp Thumb-1 findings at IDWeek 2026 this October in Washington, D.C. “This is a highly innovative and mechanistically compelling abstract with strong novelty and potential significance for antiviral drug development,” said an IDWeek 2026 expert peer reviewer. “Demonstrates conservation of an allosteric site across viral families, directly contradicting established belief that non-nucleoside antivirals cannot achieve broad-spectrum activity. This is a fundamental shift in antiviral drug development strategy.”

Model Medicines has now presented the discovery of the RdRp Thumb-1 site at IDWeek 2025^[1], AASLD 2025^[2], HepDART 2025^[3], CROI 2026^[4], ESCMID Global 2026^[5], SERVC 2026^[6], EASL 2026^[7], and ASV 2026^[8].

The MDL-001 preprint has also been featured in the recent launch of the Intercept Fund, a $500 million philanthropic initiative, backed by leaders across technology and global health, including Stripe, Anthropic, the OpenAI Foundation, Jane Street, and the Flu Lab.^[9]

The World Is No Better Prepared for the Next Pandemic Than the Last

RNA virus pandemics and epidemics occur regularly and cause widespread death and economic devastation. In the last 25 years, humanity has endured at least six major viral outbreaks: West Nile virus, SARS-CoV-1, swine-origin H1N1 influenza, MERS-CoV, Ebola, and COVID-19.^[10] The COVID-19 pandemic caused an estimated 14.8 million deaths, resulted in $13.8 trillion in cumulative economic losses, and pushed 97 million people into extreme poverty.^{[11] [12] [13]} Beyond the threat of the next pandemic, the annual toll is significant. Each winter, clinicians face a recurring tripledemic of influenza, RSV, and SARS-CoV-2 that causes hundreds of millions of infections and hundreds of thousands of deaths worldwide every year.^[14] In a non-pandemic year in the U.S. alone, adults experience two to six of these respiratory viral illnesses and children approximately six to eight, at an economic cost exceeding $100 billion.^{[15] [16]}

A Parallel Crisis in the Liver and a Different Kind of Need

Chronic viral hepatitis is a health crisis of comparable scale. Hepatitis B and C infections persist in 254 million and 50 million people worldwide, respectively.^[17] An estimated 12–72 million people also carry hepatitis D, a satellite virus that cannot replicate without hepatitis B.^[18] It is estimated that 5–15 million carry both HCV and HBV at once, yet most will never be diagnosed. Only 13% of HBV infections and 36% of HCV infections are identified globally, and fewer than 1% of co-infected patients know their status. Co-infection is not merely additive. Patients carrying both viruses are over 100-fold more likely to develop hepatocellular carcinoma than uninfected individuals.^[19] Compounding the danger, every FDA-approved HCV direct-acting antiviral (DAA) carries a black box warning for the risk of hepatitis B virus (HBV) reactivation in patients with current or prior HBV infection. This reactivation can lead to fulminant hepatitis, liver failure, or death.^[20] This FDA-mandated warning, added in 2016, applies to all DAA regimens, such as Epclusa, Harvoni, Mavyret, and others.

Together, these gaps point to a single, persistent, and largely unaddressed need: one oral agent with broad-spectrum, direct-acting activity across the viruses that drive both crises.

A Conserved Target Hidden in Plain Sight

Conventional wisdom held that allosteric sites on viral polymerases are too poorly conserved to support pharmaceutical engagement beyond genotype- or species-specific activity.^[21]  Beclabuvir, the only approved Thumb-1 inhibitor, was developed for hepatitis C and is inactive against a diverse panel of non-HCV viruses, including poliovirus, rhinovirus, coronavirus, coxsackievirus, influenza, and HIV.^[22]

The Thumb-1 pocket had gone unrecognized because it is cryptic, and Beclabuvir MoA and activity data were misinterpreted. Beclabuvir’s lack of activity across viruses led to the conclusion that the pocket either did not exist or was undruggable on other viruses. This conclusion was seemingly validated because in the polymerase’s resting state the viral Λ1-loop occupies the site, rendering it invisible to conventional virtual screening.  Model Medicines followed biology, rather than human dogma.  Model Medicines' structural biology agent identified the pocket in viruses beyond HCV. The company then analyzed published HCV inhibitor co-crystal structures to define a fifteen-residue consensus Thumb-1 pocket. Model Medicines then aligned those residues across six coronaviruses spanning both Alphacoronavirus and Betacoronavirus genera to demonstrate that the pocket and its associated Λ1-loop are conserved across RNA viral families.

Following the Science to a Novel Drug Class

Model Medicines used the molecular-geometric deep-learning model, ChemPrint^TM, within its peer-reviewed^[23], experimentally validated GALILEO™ drug discovery platform to find chemistry capable of engaging Thumb-1 across viral families. The company assembled a proprietary, three-tier bioactivity dataset spanning explicit HCV Thumb-1 chemistry and cross-family inhibitor scaffolds, then trained its model to separate HCV-restricted chemistry from genuinely broad-spectrum chemistry.

In an accompanying preprint, Model Medicines demonstrates that MDL-001 inhibits Influenza, Coronaviruses, RSV, HCV, HBV, and HDV. MDL-001 is the first Thumb-1 inhibitor demonstrated to cross viral families. These findings establish RdRp Thumb-1 as both a conserved allosteric pocket and a druggable target built to yield successive generations of broad-spectrum, pandemic-ready therapeutics.

The Discovery of Novel Druggable Targets is Rare in Virology

Prior broad-spectrum drug discovery programs focused on developing nucleoside/nucleotide therapeutics that directly target the viral polymerase active site or host-targeted antivirals. None of the compounds active against these target classes have achieved broad-spectrum FDA approvals for Universal Influenza-like-Illness (ILI) or Universal Chronic Hepatitis.

Few recently approved direct-acting antivirals act through clearly new target classes.  Notable examples include baloxavir, an influenza cap-dependent endonuclease inhibitor, and lenacapavir, an HIV capsid inhibitor.^{[24] [25]} RdRp Thumb-1 adds a new, broad-spectrum target to that short and commercially significant list.

A New Blueprint for Drug Discovery

Model Medicines broke with existing drug development paradigms to discover RdRp Thumb-1. In structure-based design, molecules are docked into a target's resolved 3D structure. This method would have failed to detect RdRp Thumb-1, because its pocket is cryptic and remains closed in the polymerase’s resting state. Conversely,  similarity-based ligand approaches rank molecules by similarity to known actives and look for close chemical relatives. This method would have failed to discover broad-spectrum RdRp Thumb-1 inhibitors because the search would have kept returning the same narrow HCV-specific chemistry. 

To discover RdRp Thumb-1, Model Medicines fused the two paradigms. To establish the biology, the team overlaid crystal structures of the HCV RdRp Thumb-1 pocket bound to inhibitors and compared the polymerase across many viral families. This comparison confirmed that the pocket is conserved. The team then ran molecular-dynamics simulations to show how a bound drug displaces the Λ1-loop. To find the chemistry, the team used its GALILEO™ AI engine to train a geometry-aware deep-learning model on a proprietary dataset of small molecules. The model learned the features that make a molecule broad-spectrum rather than the features tied to any single virus. It identified the chemistry responsible for broad-spectrum interaction with the RdRp Thumb-1 pocket.

About Model Medicines

Model Medicines is an AI-first biotechnology company engineering first-in-class small molecules that target the biological linchpins underlying disease. The company’s research spans infectious disease, oncology, and inflammation, with programs designed around conserved molecular choke points that drive multiple pathologies. Model Medicines has discovered a direct-acting, non-nucleoside, broad-spectrum antiviral (MDL-001) and a potent, selective, and novel BRD4 inhibitor (MDL-4102). Its work demonstrates how large-scale computation can uncover entirely new classes of drugs once thought unreachable. Model Medicines is advancing a new generation of therapeutics that redefine what is possible in modern drug discovery. Learn more at www.modelmedicines.com.

Media & Investor Contact

Patrick O’Neill

Head of Partnerships & Investor Relations

media@modelmedicines.com

www.modelmedicines.com