Computational Medicine

Our team does frontier research in artificial intelligence, developing mathematically sound machine learning methods which are needed for real biological and medical applications. Our goal is to develop novel accurate and computationally efficient machine learning methods for complex bioclinical challenges. We aim to integrate different types of heterogeneous data into a single model and use this integration approach to predict and explore complex human diseases. Although individual cell experiments also improve our understanding of pathologies, results from multiple datasets integrating different donors, studies and different technical platforms are essential. With regard to translational research, the data integration of different tissues and clinical conditions, is of paramount importance in identifying broad clusters of disease. Our research project includes several interrelated machine learning challenges that are the building blocks for reliable, simple, realistic, accurate, and stable models.