Variational AI is a team of AI/ML experts and business experts dedicated to using advanced generative AI for drug discovery. The company works closely with biopharma partners to develop novel therapies across disease areas, aiming to positively impact patients' lives.
Purpose Variational AI's core purpose is early-stage small molecule drug discovery. Its proprietary platform Enki™ is a generative AI platform for early drug discovery that can generate optimized, synthesizable, lead-like compounds based on user-specified target product profiles (TPP). By eliminating the hit identification (hit ID) and hit-to-lead stages, Enki™ enables pharmaceutical and biotechnology partners to move directly into the lead optimization stage and obtain novel structures that are difficult to discover with traditional methods.
Key capabilities Enki™ targets the earliest stage of small molecule drug discovery and can generate First-in-class and Best-in-class compounds. It has been pretrained across 873 drug targets, covering kinases, GPCRs, proteases, oxidoreductases, hydrolases, ion channels, and other key target classes. Unlike screening, Enki™ can generate novel molecular structures from scratch, going beyond fixed screening libraries to explore unknown chemical space. The platform supports multi-parameter optimization, enabling simultaneous optimization of more than 50 parameters, including activity, selectivity, toxicity, ADMET, pharmacokinetics (PK), synthetic accessibility, and more. Enki™ can also perform highly efficient lead optimization when the scaffold is constrained.
How to use it According to official information, the workflow consists of three steps: first, define the preclinical target product profile (TPP), specifying the molecule's target/off-target and physicochemical properties as input to Enki™'s generative AI; second, Enki™ generates novel and diverse structures that satisfy the defined TPP; third, the user selects the compounds they wish to synthesize and test. In addition, Enki™ can perform ultra-efficient lead optimization under a constrained scaffold.
Applicable scenarios Variational AI collaborates with biopharmaceutical companies and is suitable for early-stage small molecule drug discovery projects. The company has carried out collaborations with multiple organizations, for example: 3C Therapeutics uses the Enki™ platform for ligand design for its proprietary TriCore DAC platform, and this collaboration is supported by Innovate BC's Integrated Marketplace program; Variational AI and Rakovina Therapeutics are expanding their collaboration to accelerate Rakovina's kt-5000 series ATR inhibitors; Enamine has joined Variational AI's Agora open science PRMT6 discovery project, focusing on discovering novel PRMT6 inhibitors for spinal and bulbar muscular atrophy (SBMA). In addition, Variational AI has released Enki 4, a major update to its foundational model for small molecule drug discovery, with improved algorithms and architecture to support scaling, expanding pretrained target coverage from 592 to 760 targets, and is applicable to proximity-induced therapies and novel modalities.
Limitations Official materials do not provide information such as specific pricing, free trials, detailed technical documentation, or platform access methods. The platform is mainly aimed at biopharmaceutical partners and may provide services through collaboration. More usage details are not currently available.