Elix launches “Elix Discovery™,” the only all-in-one platform that provides everything needed for AI drug discovery, from models for property prediction and molecular design to AI consulting and implementation support

Elix, Inc., an AI drug discovery company with the mission of “Rethinking Drug Discovery” (CEO: Shinya Yuki/Headquarters: Tokyo; hereafter referred to as “Elix”) launched “Elix Discovery™,” the only all-in-one platform that provides everything needed for AI drug discovery, from models for property prediction and molecular design to AI consulting and implementation support.

Drug discovery using artificial intelligence (AI) and machine learning is attracting attention to shorten the time required for drug discovery, reduce associated costs, and improve success rates. However, pharmaceutical companies face major barriers in the adoption of AI due to insufficient knowledge of AI technology, difficulty in understanding how to use existing AI tools, lack of an intuitive and integrated platform, and the need to consider the best way to utilize AI for each specific drug discovery project.

Elix Discovery™ is the only all-in-one platform that provides a wide selection of AI models, intuitive Graphical User Interface (GUI), expert implementation support and consulting, and knowledge for AI drug discovery. The platform provides essential features for AI drug discovery, and will continue to grow from ongoing updates as new developments are made in the ever-evolving AI field. While many AI companies are secretive and do not disclose their knowledge and methods to clients, Elix, Inc. values transparency and openly shares knowledge with our clients to help them accumulate experience in AI drug discovery.

< Key modules of Elix Discovery™ >

1. Elix Predict (property prediction): By inputting the chemical compound structure into AI models, key information about the compound can be predicted, such as activity against a target, physicochemical properties, and ADMET properties.

2. Elix Create (molecular design): Generative AI models make it possible to algorithmically create compounds with desired properties. Molecules can be optimized for a number of factors, including target activity, physicochemical properties, ADMET, and ease of synthesis. Predictive models from Elix Predict can also be easily used for molecular optimization thanks to integrated design of Elix Discovery™.

3. Elix Assist (active learning): A suite of methods for efficiently training AI models when there is little or no available experimental data. Active learning based on predictive AI models is used to suggest the compounds that should be used to collect the next set of experimental data based on an optimization algorithm, thereby enabling the training of high performing models with fewer experiments.

< Features of Elix Discovery™ >

1. An all-in-one platform to streamline drug discovery
Pharmaceutical companies need multiple types of AI models and data processing functions to conduct AI drug discovery, and Elix Discovery™ is an all-in-one platform that implements these models and functions. From property prediction to molecular design, Elix Discovery™ has all the essential functions for AI drug discovery, and provides them in a modular architecture that allows users to leverage these functions seamlessly across the platform.

Many pharmaceutical companies feel the need to have their own AI drug discovery platform. However, building the entire system in-house entails a number of difficulties, including the many years and large costs required for development, hiring and training of talented engineers with expertise in multiple fields, and the need for on-going updates and technical support. By introducing Elix Discovery™, which has already been developed specifically for AI drug discovery and tested in the field by existing users, a high-quality environment for AI drug discovery can be established in a short period of time. Furthermore, the platform is continuously updated, allowing for the easy and rapid adoption of cutting-edge technology in a fast changing field.

2. Individualized consulting, support, and know-how provided by researchers and engineers
Pharmaceutical companies have been slow to utilize AI not only because of the lack of superior AI tools, but also a lack of know-how about AI in specific, individual situations that occur in drug discovery. In cases where pharmaceutical companies rely on consulting AI companies, it may not be possible to accumulate in-house expertise.

Elix’s researchers, who have consulting experience with numerous companies, will provide support while openly sharing their knowledge, from proposing solutions to specific issues to lectures on how to use the platform and background knowledge about AI and deep learning. For complex issues, Elix’s engineers will work on behalf of the client to create and analyze AI models, and provide them in a format that can be reproduced by the client.

Even in pharmaceutical companies that employ computational chemists, the small number of such chemists limits the support that they can provide to medicinal chemists in the use of AI. Elix provides a complete, integrated platform and consulting service for AI drug discovery, including the training of medicinal chemists who can utilize AI via the Elix Discovery™ platform.

3. Chemist-first UI/UX
Drug discovery is by no means a challenge that can be solved by AI alone; the knowledge of the medicinal chemist is critically important. Elix Discovery™ is a platform developed with this philosophy in mind, and designed to let the medicinal chemist leverage the power of AI and deep learning. Even medicinal chemists with limited knowledge of AI can use the system intuitively without worrying about choosing AI parameters and other details thanks to an automated optimization procedure when creating AI models. In addition, computational chemists who are more familiar with AI can perform detailed configurations, and easily share the models they create on the platform with medicinal chemists for immediate use. ​​

Kaken Pharmaceutical Co., Ltd. has concluded a contract to use the Elix Discovery™

Elix has already provided the Elix Discovery™ platform to Kaken Pharmaceutical Co., Ltd(President and Representative Director: Hiroyuki Horiuchi/Headquarters: Tokyo; hereinafter referred to as “Kaken”), and began consulting and supporting operations for AI drug discovery at its Drug Discovery Center (in Kyoto) in May 2022.

Kaken was established in 1948 with its roots in the Riken Foundation, which has produced many extremely important scientists in the history of Japanese science, including Yoshio Nishina, considered the father of modern physics, as well as Nobel Prize winners Hideki Yukawa and Shinichiro Tomonaga. Starting with manufacturing and sale of penicillin in the same year, Kaken has introduced many ethical drugs, including Mentax, Fibrast, and, more recently, Clenafin and Regroth, with particular strength in therapeutic areas such as dermatology and orthopedics (https://www.kaken.co.jp/english/).

Elix will continue to contribute to solving challenges in drug discovery through the on-going development of Elix Discovery™.

About Elix, Inc.
Elix, Inc. is an AI drug discovery company with the mission of “Rethinking drug discovery.” We are developing our business for pharmaceutical companies, universities, and research institutions by focusing on machine learning to reduce the enormous cost and time involved in drug discovery and improve its success rate. Visit https://www.elix-inc.com/for more details.

<Reference: Elix Discovery™ screen >

Training Detailes

Product screen (1) Model training settings

Product screen (2) Model training results

Product screen (3) Prediction results and property prediction

Product screen (4) Visualization of predictions and generated compounds

Product screen (5) Structural display of parts that contribute to the prediction (explainability)