Artificial intelligence could compress years of drug discovery into months – helping to overcome growing drug resistance to existing treatments for malaria and other vector-borne diseases. But scientists in low-income countries are often left behind. Jeremy Burrows, Medicines for Malaria Venture (MMV) vice president, head of drug discovery, explains how a new open-access, AI-powered drug discovery tool co-developed by MMV aims to level the playing field. 

Given the growing role of Artificial Intelligence (AI) in healthcare, what concrete contributions can the new technology offer in the fight against malaria?

Jeremy Burrows: At MMV, we are focused on the mission to discover, develop, and deliver new antimalarials. Along the entire value chain, from the point where you first come up with a possible molecule all the way to the end, to delivering a medicine, there are huge opportunities for AI to have an impact.

I work in the early stages of drug discovery. It is particularly in these phases where AI can play a major role in terms of helping us to identify new starting points, optimize these molecules toward candidate drugs, and then actually profile these compounds to ultimately be the medicines of the future.

AI is best at analysing large amounts of data. How do you make use of that?

Burrows: Since our foundation in 1999, MMV has worked with about two dozen pharmaceutical partners around the world, screening millions of compounds that can target the world’s deadliest and most prevalent malaria parasite, Plasmodium falciparum, which gives us extensive data. 

We’ve modelled that data and created an open-access machine learning model malaria inhibitor prediction platform (MAIP), available for free. Anyone can go there, enter a virtual chemical structure, and receive a prediction of whether that compound is likely to be active against malaria. Through validation exercises, this approach gives a tenfold increase in hit rate, which is substantial for making choices about what compounds we acquire and work on. However, our major focus is actually using AI to help optimize the medicinal chemistry design of new compounds — that is the deliberate design of compounds in a process called generative design.

So, your goal is to increase efficiency and accessibility?

Burrows: Yes, we are working with the Gates Foundation and a London-based company called deepmirror to deliver a tool called Drug Design for Global Health (referred to as dd4gh). This tool will be delivered in March 2026 and will be available for free only for scientists working in global health.

It allows a scientist to upload their chemistry and biology data, where everything is modelled. Scientists also have access to all models built on public domain data, hosted in the MMV network. With dd4gh, they can do generative design, creating virtual chemical structures that can be scored against the models to predict which ones will be good. We can boil down thousands of options to maybe 20 compounds to synthesize and test in an iterative design cycle. 

This process, called model-informed active learning, in which the models improve with each design cycle, can potentially reduce the number of cycle times required to deliver a candidate drug. Regarding accessibility, many tools like this are currently prohibitively expensive. Delivering dd4gh is therefore democratizing AI for scientists in lower- and middle-income countries to fight malaria.

You speak of democratisation. What are the main challenges of applying these AI-driven tools in low-resource settings?

Burrows: There is clear inequity in access to these technologies. The investments needed are beyond the scope of most academic groups in lower- and middle-income countries. However, there are excellent scientists in Africa, and often good internet connectivity exists – even if the data centres running the algorithms will likely not be in Africa, all the ingredients are there. 

A major challenge in drug discovery in Africa is the logistics: how long it takes to order and receive components to make a compound. We are tackling this directly by encoding DD4GH with information about the local availability of building blocks so that it would prioritize compounds to make that can be accessed quickly. Our strategy involves free access to a high-end tool, training, and co-creation together.

Does your tool rely on a proprietary algorithm? And how is data-sharing organized amongst partners?

Burrows: MMV, with the help of the Gates Foundation, has secured a licence that ensures long-term access. Critically, AI allows us to share information confidentially between partners so that it can be modelled without actually sharing data. In our recent MMV workshop where the tool was discussed, we asked participants if they would contribute their data into such a tool, and the response was unanimously positive.

What other opportunities do you see for AI in drug discovery?

Burrows: AI is opening up many new opportunities in drug discovery and MMV has multiple projects underway, including developing better models of how the immune system fights malaria and how drugs interact with the human body. We are partnering with some leaders in this field so we are excited to see what it can do.

There is widespread scepticism regarding AI. What advice do you have for drug discovery scientists facing this new reality?

The future of AI is promising but requires a healthy degree of scepticism. While there are many bold claims, it’s essential to validate new models and understand their limits. Human oversight remains critical. AI should be used as a tool to enhance, not replace, human judgment.

Continuous learning, openness to outside expertise, and adapting to new ways of working will be key. Ultimately, AI has the potential to transform how we use data, democratize technology for scientists worldwide, and co-create tools that change the way we work, especially in global health.

See related story: Looming Malaria Drug Resistance Spurs Global Search for New Treatments

Burrows spoke with Health Policy Watch on the margins of the recent MMV Science of malaria medicine symposium in Geneva.

Image Credits: Supplied by Cláudia Fançony, Violaine Martin/ MMV, DD4GH, Jeremy Burrows.