Silexon Leverages NVIDIA GPUs to Accelerate Drug Discovery Screening Process by Dozens of Times

SuiKun Intelligence is a next-generation startup at the intersection of machine learning and biotechnology, dedicated to enhancing drug R&D efficiency and success rates by ten thousand-fold through its self-developed AI platform. Currently, SuiKun Intelligence has established collaborations with nearly 20 renowned domestic and international pharmaceutical companies, listed enterprises, Grade A tertiary hospitals, government agencies, and research institutions to jointly develop new drugs and AI models.

The success rate of new drug development is extremely low, and the R&D investment is substantial. The activity, drug-likeness, and developability of drug molecules are crucial for the successful completion of clinical trials and eventual market approval. Accurately predicting these properties in the early stages of drug discovery can significantly improve the success rate of later-stage development.

The scale of big data in the pharmaceutical industry often reaches hundreds of thousands or even millions of records. The development and application of AI algorithms based on such ultra-large-scale datasets pose significant challenges to computational power. Typically, CPU computing platforms suffer from slow processing speeds; training typical deep learning models requires tens of millions of iterations and parameter updates, consuming substantial amounts of time. The application of GPUs in the field of deep learning has accelerated the development of AI. Compared with CPU computing platforms,With GPU acceleration, the time required for a single iteration of deep learning model training can be reduced from tens of minutes to just a few seconds.，Significantly improves the efficiency of AI algorithm development, shortens the development cycle, and reduces development costs.

The AI4D™ online service platform developed by Suikun Intelligence utilizes NVIDIA V100 Tensor Core GPUs. Leveraging NVIDIA GPU-accelerated computing, the platform has achieved significant improvements in computational efficiency, precision, and model training speed. Compared to CPU-only systems, its real-world application throughput has increased tenfold, substantially enhancing model training efficiency. Relative to earlier hardware generations, the integration of GPUs has delivered a tenfold increase in double-precision performance. Furthermore, by reducing the number of servers required, the platform significantly lowers data center costs.

Massive-Scale Data Processing: During the process of literature information mining, it is necessary to process over one million literature texts. With GPU acceleration, relationship extraction queries across the entire database can be completed within minutes.Speed increased by more than 10-fold, significantly improving query efficiency.

Efficient Training of Deep Learning Models:Deep learning model training speed is over 100 times faster than that of CPUs，It allows for more iterations within the same timeframe, significantly reducing time costs, effectively shortening the development cycle, and improving development efficiency.

Optimized Design of Biomacromolecules: The optimization space for biomacromolecules grows exponentially, and conventional optimization methods consume substantial computational resources. Optimization algorithms developed on GPU platforms leverage GPU acceleration and multi-GPU parallelization strategies to reduce the time required for each search iteration and increase the number of searches per unit time, enabling the design and optimization of biomacromolecules to be completed within minutes.Efficiency increased by more than 40 times。

The success of AI algorithms stems from their ability to extract deep insights from massive amounts of structured data. Suikun Intelligence leverages AI systems to integrate big pharmaceutical data, deeply empowering the entire drug R&D process. This encompasses a wide range of applications, including learning and predicting comprehensive landscapes of drug-target interactions, virtual high-throughput molecular screening, optimization of macromolecule yields, design of personalized cell therapies, and literature mining. The AI4D™ platform not only generates and screens target-directed molecules but also further predicts their drug-likeness and developability by drawing on the team’s extensive expertise in drug R&D. This approach helps focus on genuine drug candidates, thereby reducing late-stage R&D investments and improving the success rates of clinical trials and market approval.

On the AI4D™ platform, the barrier to entry for AI-driven drug molecule screening is low. Users are not required to learn algorithms or programming; they can delegate all tedious, complex, and repetitive tasks to machines, allowing them to focus their efforts on strategic research, thereby significantly reducing experimental validation costs. Traditional drug molecule screening often takes six months to a year and carries the risk of yielding no results.Leveraging NVIDIA’s computing platform to accelerate machine learning, the entire computational screening process has been accelerated by tens of times, meeting the demands of processing massive datasets, significantly reducing time costs, and thereby facilitating the advancement of new drug development.

“Machine learning is a groundbreaking tool in the field of drug discovery, holding immense application potential in biomedicine. Although machine learning does not impose high demands on computing power, fitting the high complexity of biological mechanisms with machine learning models requires more powerful computational capabilities. The NVIDIA AI computing platform brings the integration of machines and life one step further,” stated Zeng Hainian, CEO of Suikun Intelligence.