AI and machine learning: revolutionising drug discovery and transforming patient care
The development of new medicines is a complex, resource-intensive process with a high failure rate. Leveraging artificial intelligence (AI) and machine learning (ML) have the potential to revolutionize drug discovery by enhancing data analysis and prediction, leading to faster and more effective treatments.
[00:07]
In the 1970s, the biggest revolution at the time in biology — and one of the biggest in science overall — was molecular biology. The equivalent of that today is generative AI. And "lab-in-the-loop" is the mechanism by which we bring generative AI into drug discovery and development.
So, what is the lab-in-the-loop? When we try to discover drugs, we need to create new medicines — things that have never existed in the world before. And in order to do that, we bring together
[00:39]
our laboratory experiments and our machine learning and AI algorithms in an iterative process. We go back and forth between them. It’s that virtuous cycle — the integration of experimental and computational data — that will push us forward as an organization. And it will also advance the field so that we’re able to develop the medicines that will truly impact patients’ lives.
And we don’t just want to do this well — we want to do it fast. Generative AI, working together with NVIDIA,
[01:10]
gives us the ability to approach this holistically, scientifically, rigorously — but also quickly. That’s incredibly important.
I think the ability to impact people’s lives through collaboration, through working together, is tremendous. And it’s something I’m really excited about.
Too often, we assume something is not possible. But we don’t believe in intractabilities. If something truly needs to be solved, we’re going to go after it and make it a reality.
And this is a place where we want to come together to do exactly that — for the greater good,
[01:41]
for patients’ benefit, and for the future of medicines.
The process of developing new medicines is complex and resource intensive, with a high failure rate. Across the industry, approximately 90% of drug candidates fail in preclinical or clinical trials, and it can take more than ten years to determine their effectiveness. The sheer scale and complexity of the scientific data involved in drug discovery pose significant barriers to progress. Computational approaches have enhanced data collection and analysis, but have historically not matched the magnitude of this problem. Thus, there’s still potential for further advancements in the faster delivery of new medicines and improved success rates in research.
The ‘lab in a loop’ is a mechanism by which you bring generative AI to drug discovery and development.
Genentech, a member of the Roche Group, has reached an inflection point where artificial intelligence (AI) and machine learning (ML) are leveraged to redefine the drug discovery process. “The ‘lab in a loop’ is a mechanism by which you bring generative AI to drug discovery and development,” says Aviv Regev, Head of Genentech Research and Early Development (gRED). It means that data from the lab and clinic are used to train AI models and algorithms designed by their researchers, and then the trained models are used to make predictions on drug targets, therapeutic molecules and more. Those predictions are tested in the lab, generating new data that also helps retrain the models to be even more accurate. This streamlines the traditional trial-and-error approach for novel therapies and improves the performance of the models across all programmes.
The ‘lab in a loop’ strategy involves training AI models with massive quantities of data generated from lab experiments and clinical studies. These models generate predictions about disease targets and designs of potential medicines that are experimentally tested by our scientists in the lab.
By using AI approaches, we can select the most promising neoantigens (proteins generated by tumour-specific mutations) for cancer vaccines, hopefully leading to more effective treatments for individual patients. AI and ML also enable the rapid generation and testing of virtual structures for thousands of new molecules and the simulation of their interactions with therapeutic targets. AI strategies are being deployed to optimise antibody design, predict small-molecule activity, identify new antibiotic compounds and explore new disease indications for investigational therapies.
Utilising AI in drug discovery requires increasingly powerful computing capabilities to process the growing amount of data and train algorithms. In order to address this, Roche is collaborating with leading technology companies like AWS and NVIDIA. “To take advantage of these new approaches and to apply them rapidly, we need to bring together expertise from different disciplines - by doing so we have a tremendous opportunity to hopefully bring medicines to patients faster than we do today,” says John Marioni, Senior Vice President and Head of Computational Sciences at Genentech. With NVIDIA we are collaborating to enhance our proprietary ML algorithms and models using accelerated computing and software, ultimately speeding up the drug development process and improving the success rate of research and development.
“At Roche we don’t believe in impossibilities. If something really needs to be solved, we’re going to go after it and make it a reality,” adds Aviv Regev. In the next decade, the impact of AI on human health is expected to be unimaginable. AI will help untangle disease biology, predict effective approaches, and design better therapies faster, ultimately extending and improving the lives of millions of patients.