Disruption, Reconstruction, and Empowerment: The Foundational Revolution of AI for Science in Drug Discovery
PHIL RIVERS
Medical Information Technology Service Provider
Introduction
When innovative drug research and development moves from "Target-driven"Towards"Disease Understanding-Driven」, When the "Engineer Dividend" overlaps with the computing power of "Super Intelligent Agents", where is the future of new drugs in China?
In the 271st episode of PharmaCircle's "Dialogue with New Drug Leaders" live broadcast, an in-depth discussion on AI, new materials, novel modalities, and R&D philosophy may have provided answers that transcend the present moment.
Rui Zhi Medicine Hu Ruilian/Xu Yuzhou, PHIL RIVERS Zhao Yu, Hangzhou Adamerck Pharmlabs Inc. Qi Youmao, Junjian Biotechnology Xiong ChangyunFive guests deciphered the innovation code of the AI era from different dimensions.
Logical Subversion
From "Finding Targets" to "Understanding Diseases"
01
Silicon-Based Life Redefines Innovation
Yu Zhao (PHIL RIVERS): "We are not creating AI molecules; we are using AI to understand life."
Zhaoyu set the tone at the beginning: unlike the thousands of AI teams worldwide working on molecular research, their team focuses on using AI for Science to understand diseases.Biological Model。
He believes that the traditional R&D logic is "Target → Molecule", which is based on humans' limited, linear understanding of life. However, life exists in a 'high-dimensional complex space,' and the network of signaling pathways far exceeds the scope of what human experiments are capable of exploring.
PHIL RIVERS' Technological Breakthrough
New Discovery
Pathway
Independently discovered hundreds of thousands of pathways unnamed by biologists (currently, humans have discovered and named 3000 signaling pathways).
Number
Twins
Map the omics data of each individual into an activity atlas of signaling pathways.
Virtual Lin
Bed Validation
Prospective validation in collaboration with Peking University Cancer Hospital: 100% accurate prediction of drug response in 8 samples
Core Argument"The core of AI for Science is not based on human-known cognition, but to build an intelligence with an 'IQ of 1000' in the life sciences field to understand problems that humans cannot comprehend."
Zhaoyu threw out a "scare-mongering" but thought-provoking point of view:"The meaning of carbon-based life may be to introduce silicon-based life."
The host hit the nail on the head.
Wang Bo (Host)"In the past, our innovation logic was wrong. We tried to solve a disease with a single target, but the complexity of life cannot be covered by a single target. Dr. Zhao's platform explores patterns in four-dimensional space, which is revolutionary."
Platform Restructuring
New Materials, New Modalities, and New Quality Productivity
02
Qi Youmao (Hangzhou Adamerck Pharmlabs Inc.):
"The Man Behind Toxic Molecules"
As a chemist, Qi Youmao does not start from the target point, but from "Materials"Breakthrough."
Core Platform: AMCD (Cyclodextrin)
"Cyclodextrin can not only be used as an excipient but also made into a drug."
He took the core pipeline of Adamerck —— a new generation of muscle relaxant antagonist as an example:
Benchmark: Merck's Bridion (Sugammadex Sodium) with annual sales of 2 billion USD
Breakthrough: By introducing chiral structures, achieved 「Hypoallergenic"Breakthrough of "
Value: Resolved major clinical pain points of the original drug
Platformization Capability：
"We can weave countless 'cages' just like weaving a fish basket."
By grafting different amino acids and peptides onto cyclodextrin, a vast number of molecules can be designed. These "cages":
Muscle Relaxant Inclusion Compound
Capture and mask various toxic molecules (such as ADC toxins)
Alter its metabolic pathway and safely excrete it from the body.
He joked: "We are like 'fishing experts.' Just as Jiang Taigong fished for King Wen of Zhou, we fish for toxic molecules."
Xiong Changyun (Junjian Bio):
mRNA Technology "Just Entered the Emergency Room and Then the Nightclub"
Xiong Changyun pointed out the current situation of the mRNA track:
"The pandemic sent mRNA to the emergency room, and now, because of in-body CAR-T and personalized vaccines, it has entered the nightclub."
Industry Status: About 200 companies in China are already making related layouts, from Hengrui, Shiyao to startups and hospital laboratories, a hundred flowers bloom.
Core Competitiveness:"Platformization" and "Flattening Path"
"mRNA is like a copier, capable of producing thousands of proteins in the body, with the characteristics of long duration and low dosage."
AI Fusion RoadDiameter：
AI Tools: Sequence Optimization, Neoantigen Discovery
Mature CMC/CDMO System
Result: This technology has become more "flat" - even like Australia.EngineerThat way, using ChatGPT for assistance can complete the proof of concept.
Pipeline Coverage：Prophylactic Vaccines → Hematologic Tumors → Solid Tumors → In Vivo CAR-T Therapy for Autoimmune Diseases
Xu Yuzhou (RuiZhi Medicine):
AI-Powered "New Drug Engineering Department"
Xu Yuzhou reviewed the origin of Ruizhi:
"In 2002, it was established as 'Pioneer Chemistry,' with one building specifically serving Eli Lilly."ChemistryFTEProject"So it's 'Eli Lilly's first love.' " "Pioneer Chemistry"Is the predecessor of PHIL RIVERS TECHNOLOGY Co.,Ltd.
Today, Ruizhi has long surpassed chemical CROs, building:
Small molecules, large molecules, peptides, nucleic acids, ADC/XDC, PROTAC, and other full-modal discovery platforms
A Comprehensive Evaluation System from Target to IND
The Pragmatic View of AI Implementation
"AI can be a tool, a work partner, or even a higher-dimensional life form. But for CROs at this stage, the core lies in how to integrate AI tools into every aspect of production."
He divided the research and development into four stages, with AI value varying in emphasis:
Disease → Target | Early Detection: Protein Structure Prediction, Molecular Affinity Calculation |
Target → PCC/IND | Molecular Design, Drug-likeness Prediction |
IND → Clinical | Enrichment of responsive populations, optimization of clinical design |
Post-Market | Real-world Study, Indication Expansion |
Key to Data Loop Closure：Similar ClassesDeviceThe "more human-like" real and effective data generated by official platforms will be key to feeding AI and forming a closed loop, AI + organoids.
Vision："We need to turn genius-like, inspiration-like accidental discoveries into engineering inevitabilities."
Clash of Ideas
The Dialectical Relationship Between Humans, AI, and the Future
03
The most exciting part of the live stream came from the guests' impromptu debate on the role of AI.
"AI will always be a human assistant, a worker. Just like a CRO, no matter how smart it is, it still has to follow my instructions. Humans propose strategies and needs, and AI executes them. If AI becomes the master, the world will lose its meaning. Don't be intimidated by AI; we are the masters."
"I agree that AI needs humans to ask questions and make final decisions—it is a tool and decision support system. But the scale of skepticism needs to change. Faced with a constantly learning and growing superintelligence that may one day have 'a body of work' comparable to human experts, we should not cling to the authority of human experts out of habit. Its recommendations are based on exhaustive searches of high-dimensional complex spaces—something we simply cannot achieve."
"More than 50% of contamination in mRNA CMC work comes from humans (hair, gloves, etc.). AI and automation can address standardization issues and enable a 'lights-out' factory. The boundary between dry and wet labs will blur, with AI capable of making computational decisions and conducting experiments simultaneously. We are already using DeepSeek to write reports and optimize targets—AI is everywhere."
Summary by the Host
"There is no standard answer, leave it to the future. But it is enough for the tool to play its role in the present."
Conclusion
Engineering Innovation and Opportunities in China
This two-hour-long conversation ultimately outlined a clear contour of future innovation:
Four Core Conclusions
Logic Reconstruction
The starting point of innovation has shifted from "target" to "understanding of the disease."High-dimensional ComputingBecome the key to breaking through the bottleneck.
Diverse Paths
New materials (cyclodextrin), new modalities (mRNA), and new methods (organoids) are no longer single choices, but ratherCombinatorial Arsenal。
Ability Evolution
Whether Biotech or CRO, the core competitiveness shifts from "craftsmanship" to "Engineering Integration Capability", and AI is the core amplifier.
China Advantage
The enormous engineer dividend, complete industrial chain, and rapid learning and iteration capabilities have made "R&D in China, production worldwide" or even "China Defines Innovation System" has become possible.
The Final Golden Quote
Wang Bo"Why can't the Chinese redefine the ecosystem and system of innovation?"
From "Eli Lilly's First Love" to a 2.0 platform with global vision, from imitation and following to using AI to build first principles, the narrative of China's new drug development is undergoing fundamental changes.
And perhaps the starting point of all this is hidden in these "wildly imaginative" conversations.
(This article is compiled from the online live roundtable of "Yao Ding Ji × Yao Rong Quan")
Video Channel Replay Pathway
