Simcere, a Chinese Pharma Giant, Announces Global Strategic Collaboration with Schrödinger, a Pioneer in AI-Powered Drug Discovery

On July 10, Simcere (2096.HK) announced that it had entered into a global strategic collaboration agreement for drug discovery and development with Schrödinger, Inc., a U.S.-based computational drug discovery company.

Under the agreement, the two parties will collaborate on the research and development of innovative drugs focused on unmet clinical needs. Schrödinger is entitled to receive milestone payments at the discovery, development, and commercialization stages of the collaborative projects, as well as tiered royalties based on net product sales.

In terms of collaboration scope, the two parties cover different stages of the innovative drug R&D pipeline: Schrödinger focuses on the drug discovery phase, leveraging computational technologies to assist in candidate molecule design, while Simcere undertakes subsequent clinical development and commercialization.

This collaborative model of "front-end computational design + back-end clinical development" seeks to bridge computational capabilities with pharmaceutical industrialization expertise.

Schrödinger is often simplified as an "AI drug discovery company," but this label obscures the uniqueness of its technological approach.

Founded in 1990, Schrödinger is headquartered in New York, with 15 offices worldwide and approximately 900 employees. Its core platform does not rely on the "big data-driven" approach of training models on historical data; instead, it is a molecular simulation system based on physics principles—predicting the interactions between molecules and target proteins by solving the Schrödinger equation (the fundamental equation in quantum mechanics that describes the motion state of microscopic particles).

This technical approach complements AI-driven drug discovery methods that have relied on large-scale data training in recent years.

The latter typically predicts the potential activity of new molecules by learning from existing compound structures and experimental data, offering advantages in R&D scenarios with sufficient data accumulation. However, for certain novel targets with limited data and complex structures, relying solely on historical data may lead to insufficient training samples.

Furthermore, in the traditional drug discovery process, researchers typically need to test a large number of candidate compounds through experimental methods such as high-throughput screening (HTS), and then continuously optimize molecular structures based on the experimental results. This process involves extensive compound synthesis, biological activity evaluation, and iterative validation, resulting in a prolonged development cycle.

Schrödinger's computational drug discovery platform seeks to front-load certain R&D decisions by using computational simulations to evaluate molecule–target interactions prior to experimental work, thereby assisting researchers in screening for more promising candidate structures.

Simply put, traditional drug discovery often requires the experimental synthesis and screening of large numbers of compounds, followed by iterative adjustments to molecular structures based on the results; in contrast, computational drug discovery aims to narrow down the screening scope using computational methods prior to experimentation, thereby reducing ineffective exploratory efforts.

Among these, Free Energy Perturbation (FEP) is one of Schrödinger's key technological focuses. This method helps researchers predict potential changes in binding affinity resulting from different molecular modifications by simulating the impact of structural changes on binding free energy.

Compared to traditional structure-activity relationship (SAR) analysis, which requires the synthesis of dozens or even hundreds of analogs to identify promising directions, FEP can eliminate a large number of ineffective designs during the computational phase, thereby significantly narrowing the scope of wet lab validation (i.e., chemical synthesis and biological testing in actual laboratories).

After years of development, Schrödinger's computational platform has been adopted by hundreds of biotechnology companies, pharmaceutical enterprises, and research institutions worldwide. Meanwhile, the company has evolved from merely providing computational tools to actively participating in drug discovery projects, leveraging its platform to advance collaborative R&D and its proprietary therapeutic pipeline.

Currently, Schrödinger has multiple independently developed projects entering the clinical stage, and is collaborating with multinational pharmaceutical companies such as Novartis, Bristol Myers Squibb, and Eli Lilly to explore the application of computational technologies in the design and optimization of candidate molecules.

As a leading innovative pharmaceutical company in China, Simcere focuses its R&D efforts on neuroscience, oncology, autoimmune diseases, and anti-infectives, and has established the "National Key Laboratory for Neuroscience and Oncology Drug Development."

In recent years, the company has gradually established a comprehensive R&D system covering preclinical research, clinical development, regulatory submissions, and commercial operations. By the end of 2025, Simcere had successfully commercialized 10 innovative drugs while advancing more than 60 R&D pipelines across multiple therapeutic areas.

This R&D system also provides a foundational platform for external technology providers to participate in the development of innovative drugs. For computational drug discovery companies, algorithms and molecular design represent only the early stages of R&D; whether candidate molecules can ultimately become clinical assets depends on capabilities in drug evaluation, clinical development, and commercialization.

In fact, this collaboration with Schrödinger is not Simcere's first foray into AI-assisted drug R&D. In recent years, the company has partnered with domestic technology firms in areas such as AI-driven molecular screening, optimization, and candidate drug development, accumulating considerable experience in project translation.

In 2020, Simcere partnered with AlphaMab Bio to conduct virtual screening of lead compounds targeting oncology, neuroscience, and autoimmune disease areas, leveraging AI-driven machine learning models to establish target-specific screening platforms.

In 2021, the collaborative outcome ALM005, an anti-tumor molecule, entered the R&D phase. Simcere obtained the global development and commercialization rights for this molecule, marking an exploration from AI-assisted screening to the translation of innovative drug pipelines.

In 2025, Simcere further collaborated with Fermion to develop FZ002-037, an AI-driven small-molecule analgesic drug. Targeting SSTR4, the project leverages an AI platform for molecular optimization and has currently entered the clinical stage.

Multiple collaborations in AI-driven drug development have enabled Simcere to accumulate significant expertise in AI-assisted molecular screening, candidate compound optimization, and subsequent development stages. This partnership with Schrödinger builds upon Simcere's existing R&D framework by introducing a more globally capable computational drug discovery platform, further exploring the application of AI technologies in the early-stage development of innovative drugs.

The collaboration between Simcere and Schrödinger also reflects the AI drug discovery industry's transition from technological exploration to industrial validation.

Innovative drug development has long faced the challenges of high investment, high risk, and long cycles. According to Deloitte's report "Measuring the return from pharmaceutical innovation 2025," after allocating the sunk costs of R&D failures, the average R&D cost for developing a new drug among the top 20 global pharmaceutical companies in 2024 reached $2.23 billion.

In this context, AI and computational technologies are garnering increasing attention. According to the "AI in Drug Discovery Market" report published by MarketsandMarkets, the global market size for AI-driven drug discovery was approximately USD 1.86 billion in 2024, with a projected CAGR of 29.9% from 2024 to 2029, reaching USD 6.89 billion by 2029.

Market growth is also driving large pharmaceutical companies to accelerate their deployment in AI-driven drug discovery.

In 2024, Eli Lilly entered into a multi-target collaboration with Alphabet's Isomorphic Labs to explore the use of AI technologies for small-molecule drug discovery; Novartis continued its partnerships with computational drug development companies, applying AI-driven molecular simulation and generative chemistry tools to the design and optimization of candidate molecules; AstraZeneca also collaborated with firms such as BenevolentAI and Absci to explore AI applications in target discovery, molecular design, and other stages.

However, as AI-driven drug discovery gradually enters the stage of industrial application, the industry's criteria for evaluating its value are also evolving.

In the future, the value of AI-driven drug discovery technologies still needs to be validated through clinical data. For industry participants, the true focus should not be on the models themselves, but on whether AI can help companies identify more promising candidate molecules, enhance R&D efficiency, and ultimately deliver innovative drugs that meet patient needs.