DL: Towards better computational approaches and responsible innovation strategies in early drug discovery-application to antibiotics & COPD

In order to maintain the high health standards our society is accustomed to, we will need new and innovative medicines. The development of a new drug takes 15-20 years and costs on average 4 billion dollars resulting in too few new drugs every year. We believe that publicly funded research is an indispensable actor in modern drug discovery but the current technology transfer model makes translation from academia to market challenging and many promising discoveries never reach the stage where they can be taken over by private actors and benefit society. One of our aims is to propose and test a model for stakeholder involvement in early drug discovery projects, which should result in more transparency and generate socially robust knowledge. Modern drug discovery proceeds by using systematic strategies that often start with a so-called "drug target". This "target" is chosen because prior research has shown that its function or malfunction is either a direct cause of the disease or that it is involved in its symptoms. When the drug is given to the patient, it will be transported in the body and thus be able to reach the target in order to modulate its activity. The target can be seen as a lock and a drug discovery project is nothing else than the hunt for a fitting key (the drug) that can be used to unleash effects beneficial for the patient's health. During the early stages of the hunt for a new drug, it is very common to use sophisticated computer-based methods to scan through a very large number of potential drug molecules to find some fitting to the target as this will save time and money. In this project, we improve and apply dedicated drug discovery software as tools to discover new antibiotics and to find better drugs to treat Chronic Obstructive Pulmonary Disease. We are a transdisciplinary international team of scientists with expertise in chemistry, biology, mathematics, computer science, social sciences and law. In the first phase of the project, we have collected data useful for the development and testing of new computational approaches. This includes data from experiments conducted in our own labs, from public databases and from the scientific literature. Beyond being a valuable source of information for developing and testing new computational methods, the data obtained in our own labs shows that we have progressed towards our goal to propose novel drug candidates. For example we have synthesised and tested molecules that appear to be very active in relevant biological systems. In 2021, we obtained funding from the NFR (Commercialization project, Proof-of-Concept) together with Vestlandets Innovasjonsselskap (VIS) to further develop our best molecules. Our activities towards engagement of stakeholders have been frustrated by the COVID19 restrictions in 2020 and 2021, but we have progressed nonetheless and completed interviews with patients, researchers, industry, and a public interest group in 2021. We also published several articles among which "Responsible use of negative research outcomes - accelerating the discovery and development of new antibiotics" in The Journal of Antibiotics.

In order to maintain the high health standards our society is accustomed to, there is an increasing need for innovations in drug discovery. The development of a new drug takes 15-20 years and costs an average of 4 billion dollars. Modern drug discovery often starts from a validated biomolecular target for which modulators are sought using systematic strategies. In this process, an initial hit compound is optimized into a lead compound, which is further optimized to yield the final drug molecule. Computational methods are vital for the modern drug discovery pipeline and their use has the potential to accelerate and render the hit-to-lead step more cost-effective and safer. Yet, there are limitations pertaining in particular to accurate high-throughput prediction of affinity between drug targets and potential drug candidates. This creates a bottleneck where only computationally demanding methods can be applied to reach the accuracy needed for lead identification. Through this application, we aim at advancing two drug discovery projects addressing the need for (1) new antibiotics and (2) better drugs to treat Chronic Obstructive Pulmonary Disease (COPD) through the development of improved high-throughput computational methods. Further, while publicly funded research is an indispensable actor in modern drug discovery, the current technology transfer model renders translation from academia to market challenging. Stakeholder involvement is necessary for transparency and generating socially robust knowledge but not always easy to implement in such a way that the discovery project can respond to its feedback. Therefore, we also aim at proposing a model for stakeholder involvement and their influence on early drug discovery projects. To reach these ambitious goals, we have designed a truly transdisciplinary project to be tackled by an international team of experts in chemistry, biochemistry, structural biology, mathematics, computer science, social sciences and law.