HitGen Ink Deal with Cambridge Molecular to Create a Deep Learning-Driven Drug Discovery Platform

Discovering, synthesizing, and identifying hits in drug discovery is a time-consuming and expensive process. One method to speed up discovery is by using high throughput approaches such as large drug screens as DNA-encoded chemical libraries (DEL), fragment-based drug discovery (FBDD), and others. However, as the number of data increases, new approaches to analyzing the data and guiding the discovery process are necessary to reduce drug development time successfully.

More and more pharmaceutical companies are turning to Artificial Intelligence (AI), which has the ability to analyze large data sets, and it becomes better at recognizing patterns and making predictions. New types of AI, such as deep learning, go beyond that, they can take large amounts of raw data, decide what is important, and it improves over time as it learns from the outcome of its predictions.

These technologies have the potential to make drug development cheaper by accelerating the R&D phase. For this reason, in 2020, there was a record number of partnerships for AI-based drug discovery. 

With the goal of applying deep learning to DEL drug discovery, South Korea’s HitGen Inc. and UK-based Cambridge Molecular announced an exclusive alliance last week.

DEL Libraries

HitGen has established a drug discovery platform with more than 1 trillion novel small molecules. Their main screening methods include DEL libraries, FBDD, and structured-based drug design. Their business model is diverse as it ranges from screening services for other companies to developing their own therapies. 

Their pipeline currently includes three products in clinical trials:

1. HDAC inhibitor in trials for the treatment of multiple myeloma and solid tumors,
2. 2nd generation NTRK/ROS1 inhibitor, and
3. Immune-oncology agonist

Their latest collaboration with Cambridge Molecular focuses on their DEL libraries. DEL is a technique that uses DNA to tag small molecules. This makes it easier to identify the molecule that successfully binds the target protein.

Together they have created the DeepDELve 2 – Cambridge Molecular’s highly optimized DEL-specific deep learning system. According to the companies, this pipeline will provide a pathway to readily available potential ligands for drug targets. The DeepDELve2 is one of the most advanced systems currently deployed. No financial information about the deal was disclosed.

This collaboration follows last year’s acquisition of Vernalis Research, a leader in FBDD, from Ligand Pharmaceuticals for $25 million in cash and the collaboration with Evotec focusing on the development of anti-infectives.  

“We are pleased to take part in a paradigm shift in the DEL space with a partner as forward-thinking, determined, and effective as HitGen. A new capability with such an impact on the number and diversity of confirmed leads and on the speed of lead acquisition has the potential to transform the industry, enabling HitGen partners to access novel chemical space reliably and quickly, allowing faster iterations, and yielding more backup series. This is a once-in-a-generation opportunity,” said Dmitry Foxham, CEO of Cambridge Molecular.

Author

Daniel Ojeda