One of the biggest challenges in modern drug discovery is undruggable targets. These are proteins that play a critical role in diseases but finding a molecule among potential drug compounds to interact with is challenging. Emerging technologies, such as AI-driven algorithms and advanced molecular dynamics simulations, are now being used to identify hidden pockets on undruggable targets, significantly expanding the druggable landscape.

More disease-related proteins once considered “undruggable” are now being successfully targeted with small molecules. This breakthrough is opening new possibilities for curing previously untreatable diseases, including various cancers, rare genetic disorders, and other complex conditions.

It turns out that many of them have hidden pockets — sites on protein targets — that can bind to drugs. These pockets aren't visible from the protein’s structure alone but emerge when a drug binds. That is why they’re called “hidden,” “non-obvious,” or “cryptic”. Finding them enables scientists to discover compounds specifically for proteins that couldn’t be targeted before.

Accurately predicting the locations, shapes, and chemical characteristics of hidden pockets is both challenging and time-consuming. While machine learning methods are being developed to tackle this task, their effectiveness is still limited due to the small number of experimentally known cryptic pockets. There is a serious need for new methods and approaches to help find cryptic pockets more effectively.

One of Quantori partners, Ensem Therapeutics, is investing significant enterprise resources and research into solving this problem. By combining advanced computational methods with experimental AI-based technology, along with biophysics, biochemistry, structural biology, and chemistry knowledge, Ensem Therapeutics is uncovering cryptic pockets and identifying potential compounds.

The Ensem team has developed Kinetic EnsembleTM, a multi-tiered platform integrating molecular simulation, AI deep learning, and advanced experimental validation by state-of-the-art macromolecular dynamic techniques.

Research Challenge

In recent years, as the company developed and expanded, it has faced a growing demand for advanced computational methods across all stages of drug development — from target identification to compound analysis. These methods require strong computational power, such as High-Performance Computing (HPC) cloud operations, and involve complex workflows that require specialized skills, including programming and scientific computing.

One major challenge was the heavy burden on scientific developers, who had to balance operational, algorithmic, and scientific tasks. To solve this, Ensem Therapeutics needed simple, user-friendly systems that could speed up data processing, reduce errors, and improve performance. With the right tools, discovering cryptic pockets could move faster, accelerating the time-to-market for new treatments.

The company needed a partner experienced in both the domain and computational approaches. Quantori was a perfect fit with both domain expertise and knowledge of tools for chemical computations, plus extensive experience with Google Cloud Platform (GCP) services as a long-term partner.

“Quantori demonstrated strong expertise in GCP technologies and productionizing internal research tools” said Hossam Ashtawy, Director of AI/ML at Ensem Therapeutics.

Solution Provided with GCP partnership

To meet the company's scientific and technical needs, Quantori developed a flexible and cost-effective platform on GCP with data processing capabilities for molecular simulations and machine learning (ML) pipelines. This platform enhances the company's ability to run AI/ML workloads and molecular simulations at scale. Our solution includes advanced 3D visualization tools that allow researchers to quickly interpret experimental results and validate hypotheses.

Let's explore the advantages of this cloud solution and how its features could help fast-expanding companies like our partner boost their complex research:

1. Enhancing complex computations

The company’s drug development pipeline involves complex tasks, such as molecular dynamics, docking, and ML algorithms to find hidden pockets in proteins. Each task requires setting up and adjusting computations, which can be time-consuming. By using GCP Batch with GPU virtual machines (VMs), we automated these setups, allowing more calculations to run simultaneously, even under heavy load.

This not only simplified the process but also optimized resources, improved efficiency, and provided greater scalability for complex workflows, such as capturing protein dynamics and protein-ligand interactions.

2. Optimizing data operations and data systems deployment

Ensem Therapeutics wanted to improve their web application and platform. To do this, they decided to use managed services for storage and Google Kubernetes Engine (GKE) for organizing their software containers. Managed services handle the storage of data, so Ensem doesn't have to worry about setting up and maintaining storage systems. GKE helps manage and run their applications smoothly by organizing software containers, which are like small packages of their application. This approach allowed them to avoid the overhead of deploying and supporting unnecessary infrastructure. By using these services, Ensem Therapeutics made their development process more efficient and less complicated.

3. Intuitive User interface for complex computations

Quantori assisted Ensem Therapeutics in creating an intuitive web application that enables researchers to set up computational pipelines for all stages of drug discovery without programming skills. The user-friendly interface allows easy management of computational processes, saving and transferring results, and performing initial data analysis. It offers clear data representations and tools for visualizing molecular structures, helping researchers efficiently analyze data before and after computations.

Conclusion

With the high-performance, scalable infrastructure provided by GCP and Quantori expertise, our customer's scientific team is well-positioned to accelerate advancements in cryptic pockets discovery and drive future breakthroughs in drug design.

Learn more about our GCP expertise on our Cloud Services and Google Cloud pages.