US rare disease-focused biotech BioMarin has entered a preclinical collaboration with Toronto-based artificial intelligence (AI) company Deep Genomics to discover and develop oligonucleotide drug candidates for four undisclosed rare diseases with high unmet need.
According to the terms of the agreement, the partners will leverage Deep Genomics’ AI drug discovery platform, the AI Workbench, to identify and validate target mechanisms and lead candidates, while BioMarin will be responsible for pre-clinical and clinical development.
Once candidates are found, BioMarin will have the exclusive option to Deep Genomics’ rights to the programmes. In return, Deep Genomics will receive upfront and development milestone payments from BioMarin. Although BioMarin’s policy is not to disclose financial information, Deep Genomics’ CEO and founder Brendan Frey notes “the upfront and near-term payments are comparable… and similar to recent deals coming out of other AI companies, like insitro, Recursion and Atomwise”.
In September, Recursion signed an $80m deal with Bayer for AI-guided small molecule drug discovery collaboration. insitro is set to receive a $50m upfront payment and $20m in near-term payments from Bristol Myers Squibb for a partnership focused on discovering drugs for neurodegenerative diseases.
Spotlight on Deep Genomics’ AI Workbench
Founded in 2015, Deep Genomics’ work centres around the capabilities of its genetic disease-focused AI Workbench, which leverages deep learning, automation and vast amounts of in vitro and in vivo data. Frey explains the AI Workbench “is the backbone of everything we do — from identifying drug targets and designing drugs, to ensuring that the correct experimental materials and protocols are used”.
The company carried out a proof of concept in 2019 in Wilson’s disease, a recessive genetic disorder linked with the ATP7B gene that causes liver disease and neurological abnormalities, and proved the platform could “identify a gene target and design a therapeutic candidate to correct the mutation”. Frey explains: “That was a breakthrough for AI in biotech, and this program is on the way to the clinic.”
However, following this, Deep Genomics moved to expand and refine the capabilities of the Workbench. This second-generation version enables Deep Genomics to “find genetic determinants of disease, to understand disease pathology, and to identify a lead therapeutic candidate to resolve that disruption by enhancing protein expression,” according to Frey.
He adds: “For novel target mechanisms identified by our AI Workbench, 50% of them result in lead drug candidates, and we can achieve that within 12 months.” This speed and efficacy, Frey argues, is unprecedented in the AI discovery field.
Working with BioMarin on oligonucleotides
Deep Genomics’ AI Workbench 2.0 will be leveraged in its collaboration with BioMarin to discover oligonucleotide therapies. This type of therapy is a particular focus of Deep Genomics’ platform because it can successfully predict alterations in molecular phenotypes, such as transcription, splicing, translation and protein binding. Deep Genomics’ therapeutic for Wilson’s disease is an oligonucleotide and is progressing into pre-clinical trials.
Frey explains that collaborating with BioMarin is “an important proof for Deep Genomics” and its platform. “BioMarin is a [rare disease] industry leader and has its pick of AI therapeutics companies to partner with,” he says. “They chose Deep Genomics because we are unique in having technology that gives us line of sight from genetics of disease to novel therapeutic targets and drug candidates.
“This deal with BioMarin positions Deep Genomics to be the go-to company for advancing novel targets and oligonucleotide therapies.”
BioMarin chief scientific strategy officer and senior vice-president Lon Cardon noted in a statement: “We believe the combination of Deep Genomics’ experience in using artificial intelligence to creatively modulate targets coupled with our proven track record in developing transformational medicines for patients with rare diseases will speed BioMarin’s trajectory into new biological frontiers.”
Frey is hopeful that Deep Genomics will have identified a lead candidate for two of the four targeted disease indications within one year.
Deep Genomics’ broader partnership strategy
Deep Genomics is planning to use the undisclosed upfront and near-term upfront payments from BioMarin to also “extend our cash runway as we look to build our internal pipeline while also seeking other partnerships to take advantage of the broad applicability of the AI Workbench”, according to Frey.
He explains that partnering will be central to Deep Genomics’ corporate strategy as the platform is “developing programs more rapidly than any small company would be able to advance independently”. In the first quarter of 2021, Deep Genomics is expecting to sign one more pharma partnership.
Generally, Deep Genomics’ strategy is to partner with companies “interested in our current programs or a target they have been exploring”, says Frey. In addition to exploring Wilson’s disease, Deep Genomics is advancing programmes in neurodegenerative and neurodevelopment disorders, such as early-onset epilepsy.