NWO grant for smart software that searches for new medicines
Gerard van Westen and his group, together with pharmaceutical company Galapagos, start on developing software that invents new effective molecules. They will receive an NWO LIFT grant of 280,000 euros, of which 63,000 euros will come from Galapagos. The company will also bring its expertise in biology, chemistry, and computer science into the project.
Demanding
‘Developing a new drug from the start takes on average more than ten years and costs more than a billion dollars,’ says Van Westen. ‘You see, a drug is a molecule that has to meet many requirements. For example, it must be able to attach to the target in the body, but on the other hand, it must not cause any side effects. It has to have the right properties, so that it is well soluble, is absorbed into the body, and ends up in the right place.’
Contradicting demands
On top of that, different demands can contradict each other. ‘For example, a medicine that attaches well to the target in the body is often poorly soluble,’ says Van Westen.
Because of all these requirements, many drug candidates are rejected somewhere in the development process. This happens mainly in the final stages of the process, where volunteers and patients test the potential drugs, and primarily because of toxic effects or lack of efficacy. A lot of money is thus wasted on molecules that go through the whole process, but don't make it in the end. This makes the remaining medicines, of which the sale has to cover all these costs, very expensive.
The perfect balance
So the search must become more efficient, Van Westen thought. That is why he and his colleagues are now developing a smart computer program together with Galapagos. This programme invents new molecules that meet as many requirements as possible. ‘We use techniques and algorithms from computer science and therefore work together with Michael Emmerich from LIACS.’ They use a process called multi-objective optimisation. In this process, the software has to find the perfect balance: a solution that considers all the different requirements. It is the first time that this method is directly linked to programs that design new molecules.
‘This software generates new structures, inspired by existing successful structures,’ says Van Westen. ‘By means of yet other software, we can then predict where in the body these molecules will bind.’
Familiar versus innovative
When generating these new molecules, researchers must also balance between molecules that resemble existing structures and molecules that have a completely new structure. Van Westen explains: ‘Roughly put, for molecules that are only slightly different from existing molecules it is easier to predict how they will work. For completely new molecules, this activity is much less reliable.’ However, some degree of change is needed: ‘After all, we are looking for new and better medicines! In order to find molecules that are new enough, but will reasonably do what we expect, we are once again looking for the right balance. This way, hopefully, we will come to a computer program that contributes to a better and faster design process for new medicines.’
'Rather today than tomorrow'
Galapagos also regards the approach of Van Westen and his team as promising. 'Scientific innovation is in Galapagos' DNA. This collaboration fits in seamlessly with our vision on the discovery and development of innovative medicines', says Pieter Stouten, director of computational sciences at Galapagos and project manager of this project for the company. 'Artificial intelligence can help biotech companies to potentially come up with promising drug candidates faster and more efficiently, thereby helping patients and the medical community,' he adds. 'They need innovative treatment options, rather today than tomorrow.'
Galapagos supports the project financially and also contributes its knowledge and experience. Stouten: 'Our R&D team will assist with the project, but we will also offer our chemical, biological and statistical expertise, including our experience in machine learning. In addition, Galapagos will test the programme and use it in its own search for new medicines. We are proud that we can contribute to this progressive project.'
The NWO LIFT grant is financed from the ENW PPS-fund, which offers opportunities for funding of bottom-up public-private partnerships between knowledge institutions and companies within the top sectors Chemistry (incl. BBE), Energy, High-Tech Systems & Materials, Agri & Food and ICT.
LIFT stands for Launchpad for Innovative Future Technology. This specific grant is to stimulate collaboration between at least one company and one knowledge institution. The company concerned contributes part of the funding.