Model Medicines has demonstrated end-to-end drug discovery capabilities—novel target and Mechanism of Action(MoA) discovery, novel chemical discovery, and preclinical Proof-of-Concept(PoC)--against a first-in-class Target Product Profile(TPP) at unprecedented hit rates (100%) leveraging Generative-AI, Multi-Modal Models and a 0/1-Shot Discovery Approach.
La Jolla, CA — [January 28, 2025] — Model Medicines, a pioneering biotech company focused on developing first-in-class therapeutics with artificial intelligence (AI), today announced the release of its latest scientific pre-print, titled "GALILEOTM Generatively Expands Chemical Space and Achieves One-Shot Identification of a Library of Novel, Specific, Next-Generation Broad-Spectrum Antiviral Compounds at High Hit Rates.” This preprint showcases the Generative AI, New Chemical Entity (NCE), Multi-Modal Modeling, One-Shot Identification, and Library-Scale Hit-Rate capabilities of its proprietary GALILEOTM platform. Taken together with three previously released pre-prints, Model Medicines has now publicly demonstrated its end-to-end drug discovery capabilities—novel target and Mechanism of Action(MoA) discovery, novel chemical discovery, and preclinical Proof-of-Concept (PoC) against a first-in-class Target Product Profile (TPP) at unprecedented hit-rates (100%) leveraging Generative-AI, Multi-Modal Models and a 0/1-Shot Discovery Approach.
Generative AI Drug Discovery
The current paper underscores the critical need for the deployment of innovative drug discovery platforms capable of engineering precise solutions for diseases with high-unmet medical need when traditional, dogmatic approaches fail. The GALILEOTM platform uniquely enables the generation of vast chemical libraries, while discovering libraries of novel, specific and potent potential therapeutics.
Key findings from the paper include:
Generative AI: A solution space of over 50 Trillion compounds was generated and explored by the GALILEO model for this study.
New Chemical Entities: The NCE library reported in the manuscript demonstrates significant chemical novelty, with low structural similarity to both MDL-001 and known antivirals such as Beclabuvir.
Multi-Modal Models: A multi-modal discovery approach was implemented that engineered-in potency against the novel target and engineered-out a specific off-target interaction. The NCE library displayed specificity improvements of up to 15,000-fold relative to MDL-001 (see Figure).
Hit Rates & One-Shot Approach: All 12 compounds in the NCE library discovered by GALILEOTM using a One-Shot approach, from the 50T compound solution space, exhibited antiviral activity against Hepatitis C Virus (HCV) and/or human Coronavirus 229E at therapeutic relevant concentrations in vitro—a 100% hit rate.
Efficiency: GALILEOTM’s Generative AI, Multi-Modal, One-Shot approach achieved these results in a fraction of the time and cost of traditional discovery methods, highlighting its potential to revolutionize drug development pipelines.
“This research exemplifies the promise of AI-driven, end-to-end drug discovery,” said Dr. Daniel Haders, CEO of Model Medicines and corresponding author. “From target identification through to the creation of novel chemical entities to Proof-of-Concept validation, GALILEOTM integrates biology, chemistry, and computational power to engineer entirely new therapeutic opportunities with unprecedented speed and precision.”
A Multi-Generational Success Story
Building on its two prior discoveries—the identification of the novel, broadspectrum RdRp Thumb-1 target and the discovery of a novel-acting, first-in-class broad-spectrum antiviral, MDL-001, published with Icahn School of Medicine at Mount Sinai, and University of California, San Diego researchers —Model Medicines has now leveraged its proprietary, multi-modal, generative AI, 0/1 Shot platform to transform and accelerate drug discovery. The GALILEOTM platform achieved unprecedented precision in identifying first-in-class therapeutics, including, as demonstrated in the newest preprint, the next-generation of broad-spectrum antivirals.
- Novel Target Discovery: Leveraging the GALILEOTM platform’s advanced bioinformatics and multimodal AI capabilities, Model Medicines identified RdRp Thumb-1, a previously unrecognized, conserved allosteric and cryptic pocket of viral RNA-dependent RNA polymerase (RdRp), as a novel, broad-spectrum target with paradigm changing treatment potential. Read the full preprint - Discovery of RdRp Thumb-1 as a Novel Broad-SpectrumAntiviral Family of Targets and MDL-001 as a Potent Broad-Spectrum Inhibitor Thereof.
- Discovery of Novel-Acting Therapeutic, MDL-001: From this target, Model Medicines developed MDL-001, a first-in-class, best-in-class broad-spectrum antiviral with exceptional safety, ADME and potency properties. Preclinical and clinical studies in conjunction with Icahn School of Medicine at Mount Sinai, and University of California, San Diego researchers demonstrated MDL-001’s ability to inhibit numerous RNA viruses, including Coronaviridae (SARS-CoV-2, alpha-hCoV, beta-hCoV), Flaviviridae (HCV), Caliciviridae (Norovirus) and Orthomyxoviridae (H1N1, Influenza B), relative to a stringent TPP. Read the full preprint - MDL-001: An Oral, Safe and Well Tolerated Broad-Spectrum Inhibitor of Viral Polymerases.
- New Chemical Entity Discovery using Generative, Multi-Modal AI: GALILEOTM generated and explored a solution space of over 50 Trillion compounds. A Multi-Modal, One-Shot approach rapidly identified a library of 12 NCEs from this space. These NCEs, distinct from existing antiviral drugs, achieved 100% hit rates in experimental validation, demonstrating broad-spectrum antiviral activity and high specificity. Read the full preprint - GALILEO Generatively Expands Chemical Space and Achieves One-Shot Identification of a Library of Novel, Specific, Next Generation Broad-Spectrum Antiviral Compounds at High Hit Rates
- Next-Generation, Precision Engineered Pharmaceutical Solutions: Model Medicines will continue to evolve its generative-AI GALILEO platform by systematically adding complimentary discovery models, expanding our multi-modal solution space and allowing us to further precision engineering and molecular optimizations of therapeutic candidates across potency, safety and ADME variable space. Through this next phase of development, Model Medicines aims to engineer the next-generation Thumb-1-targeting NCEs through precise molecular modifications and expand the chemical solution space for this program. This engineered expansion underscores the precision of the GALILEOTM platform, which continues to evolve as a next-generation tool for transformative drug discovery against the world's most intractable diseases.
A New Era of AI-Driven Drug Discovery
In a recent paper, “Progress, Pitfalls, and Impact of AI-Driven Clinical Trials” authored by Wilczok at Duke University and Insilico Medicine’s Alex Zhavoronkov, the authors advocate for companies to develop diverse, end-to-end capabilities, rather than focus on a singular model of discovery. With this newest preprint, Model Medicines is among the few companies to demonstrate tangible results in all three AI-driven drug discovery model paradigms reported in the paper: repositioning, designing compounds for established targets, and creating entirely new chemical entities for novel targets.
Model Medicines’ GALILEOTM platform is delivering breakthroughs beyond antivirals. The company is advancing its pipeline to address unmet medical needs in oncology, rare diseases, and other therapeutic areas. With IND-enabling studies underway, Model Medicines is now moving compounds towards the clinic.
The company is also exploring partnership opportunities to expand GALILEOTM’s application, driving innovation across therapeutic domains and addressing the unmet needs of patients worldwide.
In the coming weeks, Model Medicines plans to advance the identified NCEs into preclinical studies to evaluate pharmacokinetics, safety, and efficacy across diverse viral families. Additionally, they plan to leverage GALILEOTM further to optimize lead compounds and explore additional chemical spaces for next-generation therapeutics.
VIROMME, A Model Medicines Endeavor (MEE)
The company recently announced the Model Medicines Endeavor (MME) program, a strategic approach to developing indication-specific spin-out programs to tackle urgent healthcare challenges by leveraging the company’s advanced AI-driven drug discovery platform. VIROMME is the first of Model Medicines’ indication-specific spin-outs, focusing exclusively on infectious diseases. By leveraging MDL-001 and the novel library of 12 NCEs, VIROMME aims to rapidly develop life-saving antivirals targeting the RdRp Thumb-1 site. Learn more at www.viromme.com
Preprints
MDL-001: An Oral, Safe, and Well-Tolerated Broad-Spectrum Inhibitor of Viral Polymerases
ChemPrint: An AI-Driven Framework for Enhanced Drug Discovery
About Model Medicines
Model Medicines is an AI-driven, human health company using AI to model all of chemistry and human biology, to accelerate the creation of life-changing drugs.
The company was founded in 2019 to deliver on the promise of AI-Drug discovery. They have discovered 192 compounds and advanced 67 assets in cellular models of disease, or beyond, across 12 therapeutic targets for multiple areas of biology. Their data has been validated by preeminent researchers and scientists at premier academic and corporate laboratories. The company has developed a robust pipeline of patent-pending therapeutics for oncology, infectious diseases, gastric disorders, neurological disorders, and weight disorders. The company is based in La Jolla, CA.
To learn more, visit www.modelmedicines.com
Figure - Comparative IC50 Analysis of Thumb-1-Targeting NCEs and MDL-001
Figure: Comparative IC50 Analysis of Thumb-1-Targeting NCEs and MDL-001. The plot displays half-maximal inhibitory concentration (IC50) values for the 12 NCEs and MDL-001 in MCF7 growth inhibition assays. The left y-axis represents IC50 values, while the right y-axis shows relative IC50 compared to MDL-001.
