Galixir Announces PCC Discovery for Autoimmune Disease, Marking a New Era of Technical Validation in AI-Driven Drug Discovery

Today, Galixir announced the utilization of its self-developed artificial intelligence drug discovery platform, Pyxir.^®A novel preclinical candidate compound (PCC) for the treatment of autoimmune diseases has been identified, validating its AI-driven, technology-powered drug discovery capabilities in a closed loop.

Galixir, an AI-driven small molecule pharmaceutical company founded in 2019, has completed five rounds of financing.

"Once the technical closed-loop is completed, achievements such as reaching milestones and securing pharmaceutical company partnerships will rapidly increase." Li Chengtao, founder and CEO of Galixir, told VCBeat. "I believe that for any company, a key moment like this in achieving a technical closed-loop signifies when it is ready to take off."

AI in the medical field has never lacked disruptive, game-changing myths, but entrepreneurs like Chengtao Li have become increasingly aware,AI Drug Discovery is a Gradual Revolution, and More Importantly, a Result-Oriented Industry. The Most Important Thing for Companies is to Demonstrate Their Technical Strength.With the development of AI technologies such as AlphaFold2, AI pharmaceutical companies have gradually completed their technical accumulation and platform construction, bringing AI drug discovery to a new stage of process and technical validation.

Unlike the backgrounds of most founders of AI pharmaceutical companies, Li Chengtao comes from a computer AI background. In 2014, Li Chengtao graduated as an outstanding graduate from Yao Class at Tsinghua University, and in early 2019, he obtained his Ph.D. in Computer Science from MIT. In the same year, he founded Galixir.

MIT and Boston have a strong entrepreneurial atmosphere, with many startup organizations. Li Chengtao was deeply influenced by this environment and has become a representative of the new generation of entrepreneurs: they had entrepreneurial aspirations at an early age, with clear goals, and are on the same starting line as entrepreneurs worldwide in the same field.

Unlike the founder backgrounds in other subfields of the pharmaceutical industry, many founders of AI drug discovery companies come from technical backgrounds. However, the AI drug discovery sector is an interdisciplinary field that encompasses computer science, biology, chemistry, and more. As an increasing number of AI drug discovery companies shift from the CRO model to the Biotech model, the demand for multidisciplinary talent in the industry is growing.

Although Li Chengtao focuses on the AI field, he pays great attention to his learning and accumulation of pharmaceutical knowledge. The team members cover multiple fields, not only having experience in advancing several drug research and development projects to clinical trials but also possessing machine learning implementation experience from top internet companies such as Google, Apple, and Microsoft.

"In the past, when few people knew about the AI pharmaceuticals field, almost no one ventured into this direction. Now, universities like Tsinghua, Peking, Fudan, Zhejiang, and Shanghai Jiao Tong are actually launching projects or talent cultivation programs in AI pharmaceuticals." Li Chengtao believes that as awareness of the AI pharmaceuticals sector grows, coupled with the gradual expansion of systematic education, there will be an increasing number of interdisciplinary talents.

Before starting his business, Li Chengtao did a lot of preparatory work related to entrepreneurship, not only accumulating knowledge, technology, practical experience, and entrepreneurial connections, but also, as the chairman of MIT-CHIEF (MIT Chinese Innovation and Entrepreneurship Forum) in 2018, he frequently traveled between the US and China due to his responsibility for the Boston entrepreneurial team's China trip. This allowed him to further understand the broader industry environment in China.

To gain a deeper understanding of the latest advances in drug research and development, as well as the industry insiders' perspectives on AI, Chengtao Li once spent three months reaching out to and communicating with nearly 100 professionals through LinkedIn. "After talking with them, I realized that, in fact, the majority of people, whether they are in drug development, AI, or investment, do not actually approve of this approach."

Li Chengtao realized that the lack of recognition was mainly due to two reasons. On one hand, it was because of the lack of understanding of the interdisciplinary knowledge of AI and drug development; on the other hand, it was due to the fact that AI technology had not made a breakthrough at that time.Many AI companies have developed for a long time and raised a lot of money. But in fact, their algorithms cannot be implemented, let alone make money.

The pharmaceutical industry has undergone multiple technological revolutions, with many technologies coming and going. Very few technologies that can truly generate value remain. Before AI "revolutionizes" the pharmaceutical industry, the most realistic problem at hand is still pushing drugs into clinical trials.

It will take time for people to cultivate and improve their understanding of AI drug development. This was "the biggest difficulty" Chengtao Li faced in the early stages of his entrepreneurship.

To change cognition requires cutting-edge technological breakthroughs. In 2018 and 2020, the two generations of AlphaFold algorithms solved the 50-year-old problem of protein spatial structure prediction in the biology community, directly igniting the entire industry, and the potential of AI pharmaceuticals began to truly emerge.

However, the entire industry is result-oriented. No matter how impressive the technology sounds, the result is the ultimate criterion.And technical strength can only be reflected through each pipeline."From the initial target selection, to design, optimization, screening, synthesis, testing, and finally iterating to produce preclinical candidate drugs, this process is actually the best proof of one's own technology and the entire industry's technical capabilities."

In the process of collaborating with multinational pharmaceutical companies and large domestic pharmaceutical enterprises, Li Chengtao learned about the real needs of big pharmaceutical companies for AI-driven drug discovery companies: First, whether AI-driven drug discovery companies can do what pharmaceutical companies can do and do it better; second, whether AI-driven drug discovery companies can achieve what is difficult for pharmaceutical companies to accomplish, such as rational design of small molecule candidate drugs for hard-to-drug and undruggable targets.

Galixir's latest preclinical compounds, generated, screened, and iterated by AI, not only feature a novel scaffold structure and independent IP space but also demonstrate excellent performance in efficacy, pharmacokinetics, and safety. They have shown promising results in both in vitro and in vivo preclinical studies.

It can be said that the AI pharmaceuticals industry is transitioning from a technological revolution to a phase of technical validation, with an increasing number of AI pharmaceutical companies introducing candidate compounds, and some even advancing their candidate pipelines into clinical trials.According to the "QuantumBit AI Pharmaceutical In-Depth Industry Report," before 2021, there were only 3 AI pharmaceutical pipeline products globally that had publicly entered the clinical stage; this number increased to 14 in 2021.

Galixir is advancing more than ten drug research and development pipeline projects. Meanwhile, it has also recently disclosed the latest R&D progress.

After algorithms like AlphaFold solved the protein structure prediction problem, the prediction of molecular and protein interactions has become the final piece of the puzzle in AI-powered drug design.If this problem can be solved, the entire pathway will be打通 (dǎtōng), linking target selection, structural analysis, protein feature extraction, rational design, and other stages to achieve a technological closed loop in drug design.

In April, after Galixir and AstraZeneca Global R&D Center jointly released the structure-based deep affinity prediction model STAMP-DPI, they collaborated with researchers from Fudan University and Sun Yat-sen University to release the latest TBind v1.0.1, which is based on the Trigonometry Aware Neural NetworK (TANK) and focuses on 3D structure prediction of small molecule ligand-protein complexes.

"This is the world's first deep representation learning framework capable of simultaneously predicting the three-dimensional binding conformation and binding affinity of small molecules and target proteins, significantly surpassing the results of existing methods—especially those that have been industry standards for many years," said Li Chengtao. "We are very excited to complete the final piece of the puzzle from our side. I believe this is a crucial milestone for the industry."

From the perspective of the entire industry, AI pharmaceuticals are gradually gaining recognition. According to incomplete statistics from VCBeat, in 2021, there were 34 financing events in the AI pharmaceuticals field in China, involving a total financing amount of 8.3 billion yuan.

Not only are investors willing to pour money into the sector, but pharmaceutical companies are also becoming increasingly open to collaborating with AI drug discovery firms. Numerous partnerships have been formed, such as Sanofi and Exscientia, Hengrui Medicine and Iktos from France, Fosun Pharma and Insilico Medicine. There have even emerged large AI drug discovery alliances led by traditional pharmaceutical enterprises and universities, such as MELLODDY and MLDPS.

When will the AI pharmaceuticals industry enter its explosive growth phase? Li Chengtao provides his answer:"I think it's almost time. After completing the first round of technological revolution, we are now ushering in the second round of technical verification. Once the entire process and technology have been validated, it will be time for scaled expansion."

Galixir has just completed the puzzle of drug design, and its technology has formed a closed loop. In the short term, Li Chengtao hopes to secure more PCC projects and aims to establish deeper collaborations with more pharmaceutical companies both in and outside China.

In the long term, he hopes to challenge more challenging and innovative targets in the future, especially the development of undruggable and difficult-to-drug targets. In fact, Galixir has already initiated related research, including recent dual-target PCC and intrinsically disordered proteins (IDPs) targets, achieving promising results.