deepmirror: Generative Drug Design at scale
Welcome back to the “Deep Dive“ series where we bring to you some cool AI tools that will have a strong impact in Life Science. In each episode, we bring an exclusive interview with the main developers behind cool tools and we tell why the tool is important, what it does and what the future holds!
Today we present deepmirror, an online platform for generative drug design that recently raised $2.4M! We have interviewed the co-founder, Andrea Dimitracopoulos, who shed some light on why deepmirror was created!
🔴 The Problem
In the field of drug discovery, there’s no shortage of AI models and open-source code, but many of these tools fail to ultimately help scientists because they are difficult to implement. For non-technical teams, setting up local environments, troubleshooting code, and managing complex dependencies can be frustrating and time-consuming, leading to many valuable tools being untouched.
Scientists often prefer user-friendly interfaces where they can focus on their research rather than technical hurdles. AI holds the promise of learning from experimental data, speeding up drug discovery by predicting better molecular candidates and optimising properties like potency and ADMET. Yet, without accessible platforms that simplify this process, many teams are unable to fully harness the power of AI in their drug programmes.
💡The Idea
deepmirror was created to bridge this gap between cutting-edge AI tools and the practical needs of biotech scientists. The idea is simple: instead of downloading code and running complex scripts locally, deepmirror offers a seamless web-based platform where users can easily apply AI models to their drug discovery efforts without any technical barriers.
deepmirror’s platform is designed to learn from experimental data, helping chemists refine molecular designs and predict drug properties like potency and ADMET with greater accuracy. Scientists can upload their own datasets and let deepmirror’s AutoML capabilities do the heavy lifting. The platform runs hundreds of AI models in parallel, selecting the best-performing one for the task, making state-of-the-art machine learning tools accessible to a broader range of scientists.
What sets deepmirror apart are its generative capabilities tailored specifically for chemistry. This AI model generates new molecules based on project goals, giving scientists the ability to explore novel chemical structures that align with their experimental needs, and iterate more rapidly. For medicinal chemists, this represents a meaningful leap forward: it shortens the idea-to-experiment timeline and ultimately could help accelerate the journey from concept to patient-ready treatments.
🔮The Future
deepmirror aims to evolve into the digital hub where all therapeutic development and co-ideation happens, from the initial idea all the way to clinical trials. 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.