Bioinformatics researcher Katharina Waury shows how bioinformatics can support the process of biomarker development. This allows scientists to conduct more targeted research into dementia.
Waury's research focused on improving biomarker assay development for dementias such as Alzheimer’s disease. Biomarkers are biological measurements that provide information about a person’s health state. For example, within the dementia field, researchers try to identify proteins within the blood or other body fluids that can indicate if and when a person will develop Alzheimer’s disease.
Finding and developing novel biomarkers is often unsuccessful. One major reason for this is the difficulty in creating biomarkers assays. These assays use antibodies which can bind and thus measure the biomarker. Selecting the right antibodies for a new assay is very important but difficult to do. Researchers often use a trial-and-error approach which requires a lot of time, work and materials. Waury therefore explored how computational methods can improve biomarker development by analyzing large sets of protein data and predicting which biomarkers and antibodies are most likely to work well in clinical assays.
Her research showed that using bioinformatics can support the biomarker development process. Online tools and biological datasets that are freely available online can help researchers to choose the best biomarker candidates or identify potential issues with a candidate early on. Further, Waury and her colleague developed machine learning (or AI) models that can predict properties of a protein that are relevant to decide if a protein is a good biomarker candidate. They have also investigated which parts of a protein are most suited for an antibody to bind as is necessary in biomarker assays. With this multifaceted approach, they are providing tools and insights to biomarker researchers to help them develop novel biomarkers for clinical use.