Variational AI

Variational AI is a pioneering company at the intersection of artificial intelligence and biotechnology, dedicated to accelerating the drug discovery process. Founded in 2019, the company has developed Enki™, the first commercially accessible foundation model specifically for designing novel small molecules. This powerful generative AI platform empowers biopharmaceutical companies to move from target identification to preclinical testing with unprecedented speed and efficiency, fundamentally changing the cost and time dynamics of developing new therapeutics.

Instead of screening millions of existing compounds, Variational AI's approach is to create new ones from scratch. The Enki™ platform is trained on decades of experimental data, encompassing over 570 targets across more than 10 major target classes, including GPCRs, kinases, proteases, and NHRs. This extensive training allows the model to understand the complex rules of molecular interactions and chemistry, enabling it to generate molecules with specific, desired properties.

How to use Variational AI

The workflow with the Enki™ platform is designed for simplicity and precision, allowing medicinal and computational chemists to translate their strategic goals directly into molecular candidates.

1. Define the Target Product Profile (TPP): The process begins with the user defining a clear TPP. This is the most critical step, where the desired characteristics of the final drug candidate are specified.
2. Select Targets and Off-Targets: Users specify the primary biological target(s) the molecule should interact with (on-targets) and, just as importantly, those it should avoid (off-targets) to minimize side effects and increase selectivity.
3. Specify Physicochemical Properties: Beyond target-binding, users can define other crucial properties such as solubility, permeability, metabolic stability, and even brain-penetrance for CNS-targeting drugs.
4. Generate and Prioritize: Once the TPP is defined, Enki™ gets to work. The generative algorithm explores the vast chemical space to design novel molecules that fit the multi-property profile. It then provides a prioritized list of the most promising candidates, significantly reducing the number of compounds that need to be synthesized and tested.

Core Features of Variational AI

• Generative Foundation Model (Enki™): A state-of-the-art generative algorithm that creates, rather than just searches for, novel small molecules.
• Extensive Training Data: Trained on a massive dataset covering over 570 targets and 10+ target classes, ensuring broad applicability.
• Data-Independent Operation: A key advantage is its ability to generate molecules for novel targets where little to no experimental data exists.
• Multi-Property Optimization: Simultaneously optimizes for on-target potency, off-target selectivity, and desirable ADME (absorption, distribution, metabolism, and excretion) properties.
• Rapid Lead Generation: Capable of delivering novel and selective lead structures in a matter of weeks, a process that traditionally takes many months or years.
• High-Efficiency Discovery: Proven to be highly efficient, with case studies showing that a small set of AI-generated molecules can be more effective than screening a million compounds via traditional methods.

Use Cases for Variational AI

Variational AI collaborates with biopharmaceutical partners across various therapeutic areas to tackle challenging drug discovery programs.

• Oncology: In partnership with Rakovina Therapeutics, Variational AI used Enki™ to accelerate the discovery of a brain-penetrant ATR inhibitor, a challenging objective for treating cancers that have metastasized to the brain.
• Immunology & Inflammation: The company is co-developing a novel small molecule inhibitor for chronic skin inflammation, specifically atopic dermatitis, in collaboration with Oncocross.
• Immuno-Oncology: ImmVue Therapeutics selected Variational AI's platform to discover first-in-class cancer drugs, leveraging Enki™ to generate novel molecules for immuno-oncology targets.

Advantages of Variational AI

The platform offers a paradigm shift from traditional drug discovery methods, providing significant competitive advantages.

• Speed: Drastically reduces the timeline from hit identification to lead optimization.
• Cost-Effectiveness: By minimizing the need for expensive and time-consuming high-throughput screening (HTS), it lowers the overall cost of early-stage discovery.
• Innovation: Generates truly novel chemical matter, opening up new intellectual property opportunities and pathways to drug previously 'undruggable' targets.
• Higher Success Rate: By designing molecules with better properties from the outset, it increases the probability of candidates succeeding in later preclinical and clinical stages.

Pricing and Plans

Variational AI operates on a partnership and collaboration model. Pricing is not publicly listed and is customized based on the scope of the project, the therapeutic area, and the nature of the collaboration. This typically involves co-development agreements, licensing of the Enki™ platform, or milestone-based payments. Interested biopharmaceutical companies are encouraged to contact the company directly to discuss potential partnerships.