Bioinformatics

Research interests

The aim of the Bioinformatics group is to study and understand molecular mechanisms of human immunity related processes and diseases utilizing the tools of computational and systems biology.

The Group has two major research lines. Computational immunology studies aim to elucidate the system level working of molecular components of human immunity. We have identified genes and proteins forming the human immunome. Collecting information about these genes and proteins allow us to find relationship between distant datasets which allow system wide analysis of immunity. We have investigated and simulated certain immunological processes based on immunome interactome (protein-protein-interactions) and gene expression information. These studies have revealed new features of biological networks and allowed e.g. successful prediction of novel primary immunodeficiency (PID) candidate genes. Another approach to investigating the immune system at system level is to follow gene and protein expression during B-cell development by applying microarray and proteomics methods. We are now applying systems immunological approaches to study the mechanisms of PIDs to learn how these diseases could be efficiently treated on system level.

The other research line focuses on the analysis and prediction of the effects of variations ranges from collecting information into publicly accessible databases to detailed structural and functional studies of the effects and consequences of the variations. Currently we maintain about 140 variation databases. These databases mainly contain information on immunodeficiencies and cancers. The effects and consequences of the variations are investigated at RNA, DNA and protein level. The ultimate goal is to help in developing new therapies to treat patients. We are developing tools for the prediction of numerous effects on protein structure, function, interactions, stability, aggregation and disorder propensity etc. For this purpose we provide the Pathogenic-Or-Not -Pipeline (PON-P), that allows easy and fast access to a large number of predictors and interpretation of the predictions. We have successfully predicted the effects of thousands of variations. We have also investigated the performance of numerous prediction methods with large test sets. For these studies we have collected benchmark datasets in VariBench, a suite for high-quality experimentally verified variation information.

In addition to research, the Bioinformatics group is in charge for the organization of the curriculum of the international Master's Degree Programme in Bioinformatics at the University of Tampere.