Deep Genomics Files IPO Prospectus, Pioneering AI-Driven Drug Discovery Platform

Deep Genomics is a Canadian company that leverages machine learning to track potential genetic factors. VCBeat has learned that on May 3, the company officially announced the application of its technology to drug development. Their technology can help produce powerful new drugs by detecting subtle signals within large-scale genomic data.

Deep Genomics, founded by University of Toronto professor Brendan Frey, specializes in machine learning and genomics. The company uses deep learning, or very large neural networks, to analyze genomic data. By identifying one or more genes responsible for a disease, it helps researchers develop drugs that can correct aberrant gene behavior.

To date, the company has focused on researching hard-to-detect genomic mutations that may be associated with specific diseases. The company will initially prioritize early-stage drug development for Mendelian disorders, which are genetic conditions caused by single-gene mutations. Globally, 350 million people are affected by these disorders.

The primary reason for applying AI technology to drug development is the emergence of powerful new algorithms, which also enable whole-genome sequencing by reducing costs and improving efficiency. “We are entering a new era of data-rich, information-driven medicine,” said Frey. “You can access many different types of data, and the best technologies we have for processing large volumes of data are machine learning and artificial intelligence.”

In recent years, deep learning has become a powerful tool for identifying abstract patterns using large volumes of training data and has proven particularly valuable in speech recognition and classification. This approach also offers a method for detecting diseases in medical images and has enabled the prediction of diseases from patient records.

Frey, a former computer scientist who pursued advanced studies at the University of Toronto, stated that Deep Genomics will seek collaborations with pharmaceutical companies for drug development. Deep Genomics primarily provides expertise.

Frey stated, “In about five years, pharmaceutical companies that want to succeed must use these AI tools.”

Deep Genomics also specializes in uncovering elusive and less direct disease triggers within the human genome. For example, the company has published findings demonstrating how deep learning can help identify genes in DNA that contribute to diseases such as spinal muscular atrophy and non-polyposis colorectal cancer.

Stephen Saunders, an assistant professor at the UCSF School of Medicine in San Francisco, who specializes in using genomics and bioinformatics to study diseases, stated that deep learning can facilitate drug development by identifying opportunities within sparse pathological data and large-scale genomic datasets.

In addition to Deep Genomics, several other companies are also applying machine learning to drug development, including the UK-based company BenevolentAI and Calico, a subsidiary of Alphabet.

Ken Mulvany, founder of BenevolentAI, stated that his company focuses on inflammatory and neurodegenerative diseases, orphan diseases, and rare cancers by leveraging largely underutilized research data. Mulvany remarked, “Drug development remains a protracted, risky, and costly process with high attrition rates. However, there is a vast amount of untapped data within the R&D departments of pharmaceutical companies that have no development plans for it.”