– Contributing to innovation in new drug development by shortening the period for deriving new drug candidates

– Applied for a patent to discover immune-active substances with COVID treatment effect by using AI model

Currently, the world economy is in turmoil due to the prolonged COVID-19 outbreak. However, IT, content, and bio companies located in Pangyo did well in domestic and overseas markets, and this gives hope to related industries and many startups. In this issue, we looked into the achievements and major activities of AZothBio that developed 10 models such as deep learning-based drug activity prediction and virtual search/generation, thus innovating the new drug development process.

AZothBio has developed 10 models such as deep-learning-based drug activity prediction and virtual search/generation over three years since its establishment in 2016 by its own technology, and utilized AI platform to jointly carry out seven new drug exploration and development projects with partners. “Conventional method of developing new drugs cost so much time and money and also too risky to perform in Korea because it had a high possibility of failing,” said Shin Jae-min, the CEO of AZothBio. He also added, “As the environment that shares AI technology’s development and big data, we were able to innovate the process of new drug development, and that is why I founded this company, to contribute to the development of the medical industry.”

The major AI platform developed by AZothBio is AiDL®, which is an integrated platform through which AI models for new drug development can operate efficiently. Shin explained, “AiDL® is a platform that enables the development and exploration of Ai-based new drugs and activity prediction to operate in integration. This drastically reduces the delivery of a new drug candidate which usually took 4-6 years by traditional method.” It is AZothBio’s goal to reduce the delivery period of a new drug candidate with AiDL® to 1.5 to 2 years. Dl Company has built functions that are necessary for the discovery of new drug candidates such as virtual search, binding force prediction, activity prediction, off-target prediction, liver toxicity prediction, and creation of new substances in AiDL®. And, it is working with pharmaceutical companies in Korea and overseas to discover new drug candidates.

Also, this company is developing diverse models like AiKpro, which is used for target screening of the drug and predicting off-targets, AiCAD/CAD-S that can be referenced for the direction of research and developing indications as the model that predicts drug activity compared to 980 cancer cell lines, AiP450 that predicts type CYP5 that plays a major role in the metabolism (lyase) among drugs, and AiTope that can be utilized in the development of immune activator that can correspond to the anticancer vaccine or infectious diseases.

Shin said, “The characteristic of our AI model is that it has been built as a data platform that can process, analyze, and utilize heterogeneous and diverse types of data. With the platform called AiDL®, we explored new substances, predicted the drug efficacy, and organically operated the necessary process for exploring and developing such as virtual creation and synthesis. AiMol-ensemble® creates new substances by virtually combining the characteristics of the substances to create new substances, so we can quickly respond to the needs of the pharmaceutical companies.”

AZothBio has recently applied for domestic and international patents for a peptide substance that induces immune activity by using its own AI model, which is effective in treating COVID. The immunological activity-inducing substances discovered this time are peptide-based compounds, which were discovered using the nucleotide sequence information of COVID. The initial verification of AZothBio’s epitope prediction model AiTope was made by completing the prediction accuracy test at an experimental agency in the UK. The immunological activity prediction verification of the new peptide substance for COVID and the patent-pending was conducted with a partner company, Qingdao Kezishing Ltd. AZothBio has the right to domestic and overseas businesses excluding China, and is discussing development and commercialization with specialized biotech.

(From left, Managing Director Maeng Jae-yeol and Team Leader Park Hae-jin of AZothBio)

AZothBio has successfully attracted investments from various organizations, including national support projects such as AI-based drug development platforms. It is now based in Pangyo, the place where many headquarters of bio venture companies and pharmaceutical research centers are located, and is actively expanding its business. “We are a platform company using AI, and therefore have few restrictions on entering the region or market,” said Shin. He plans to expand joint research with various domestic and foreign drug development companies, pharmaceutical companies, and relevant organizations.

In particular, the company plans to secure leading materials by next year by “choosing and focusing” on projects with high potential among seven projects currently underway. To make stable profits, the company is getting ready to launch the platform service and do an early-stage license out on the candidate by 2022. Also, the company plans to expand business channels by using cloud platforms.