deepmirror secures $2.4 million to help chemists design drugs with one click
Bringing a single drug to patients may cost more than $1 billion – roughly as much as building and launching a spaceship. Up to half is spent on optimizing drug molecules for safety and efficacy, a process that can take up to 6 years and delays drugs from reaching patients. While AI could help chemists design optimal ones faster, implementing it effectively remained out of reach for all but the largest pharma companies. Today, deepmirror has raised $2.4 million to simplify the adoption of AI for molecule design and enable every company to get drugs to patients faster with their accessible AI-powered platform. The platform empowers chemists in using data to generate optimal molecule ideas with a few clicks of a button – giving them access to best-in-class AI capabilities without building in-house or having to partner with other companies.
deepmirror's $2.4 million Seed round is led by London-based Twinpath Ventures and matched with an Innovate UK Investor Partnership grant, with additional participation by Saras Capital, US-based stealth.vc, and experienced business angels from leading biotechnology firms.
Today, medicinal chemists spend years refining molecules to ensure they are potent enough to treat a disease and safe enough for human consumption. The process is comparable to solving a Rubik's cube with a hundred different colors as drug potency, selectivity, oral availability, toxicity, metabolic stability and more must be balanced against each other. Often, improving one of those properties worsens another one and chemists thus frequently make and test over 1,000 molecules per drug program to find the sweet spot. deepmirror's platform eliminates this trial-and-error by bringing together data from previous experiments and a chemist's intuition to only suggest the most promising molecules for testing.
UK-based deepmirror was spun out from the University of Cambridge in 2019 by Dr Max Jakobs, Dr Andrea Dimitracopoulos, and Dr Ryan Greenhalgh to improve the adoption of AI in the Life Sciences. During their career as scientists, the team realized that the biggest barrier to AI adoption in drug discovery is not availability of great modelling approaches but their accessibility to chemists. Hence, the trio collaborated with experienced chemists to design, build, and launch the deepmirror platform in 2023.
deepmirror's AI engine combines cutting edge generative and predictive AI to generate only relevant molecules with flawless user experiences to make AI work for chemists. Notably, deepmirror takes desired ranges of molecular properties and other user input into account in the generative process. Furthermore, deepmirror's predictive AI Engine utilizes meta learning and multi-model selection to achieve state-of-the-art performance on many molecular property prediction tasks, outperforming other industry solutions.
The company is currently making the platform available to chemists in small to medium sized biotech businesses – and the impact of the deepmirror platform is already visible in real-world applications. For the Medicines for Malaria Venture, a non-profit for malaria research, it identified antimalarials with 10x less off target activity in just one hour. The platform also accelerated 3 months of drug discovery down to just 2 weeks for a European biotech who needed support identifying first-in-class oncology drug candidates with better Absorption, Distribution, Metabolism and Excretion and Toxicology (ADMET) properties. The company used deepmirror to parse their chemical library — nearly 5000 compounds — to rapidly identify the top 10 and now routinely screen millions of compounds on a regular basis with programmatic access to deepmirror.
Dr Max Jakobs, Co-founder and CEO of deepmirror said:
deepmirror was founded to give chemists the power to design drugs faster and more efficiently without the need to build AI infrastructure internally or sacrificing IP to partners. This funding will help us continue making AI accessible so that any team, no matter the size, can innovate and accelerate drug discovery on their own terms.
Although the broader biopharma industry is increasingly embracing AI, it’s hard to build in-house. There are too many models to choose from and partnering with AI companies often requires relinquishing IP. deepmirror is committed to making AI accessible and easy to use without the burden of partnerships and substantial investments. With immediate, impactful results for chemists aiming to accelerate drug development.
We used the deepmirror platform on an active drug discovery program to predict the experimental biological activity outcome of our whole proprietary library and identified and validated novel active scaffolds
Francesco Greco, Senior Research Scientist, TES Pharma:
deepmirror's seed funding will be directed towards doubling down on the platform’s capabilities. While currently focused on optimizing small molecules, the company’s vision is to broaden its scope so that they can support the design of all molecule modalities, including peptides and antibodies. Eventually, deepmirror hopes to be the digital hub where all therapeutic development and co-ideation happens—from the initial idea all the way to clinical trials.
John Spindler, Partner, Twinpath Ventures:
deepmirror will transform the drug discovery landscape by giving chemists the tools they need to leverage AI without the typical barriers. We believe their simple to use AI platform is a game-changer for biopharma companies of all sizes, offering a faster, smarter way to develop new drugs. We're excited to lead this investment round in deepmirror and fully expect Max and the team to continue to push the boundaries of what’s possible in drug design.
