Home | ResearchCV | Publications | Personal |   
Immunoinformatics

Every foreign agent that enters the body has the potential to elicit and immune response. Not only pathogens, but also every drug, every food, or even cosmetics can potentially generate unwanted immune responses. Nobody wants to be allergic to their shampoo or favorite cereal.

Our immune system usually, keeps us disease-free, as ‘self’ is tolerated and ‘non-self’ attacked and eliminated. When ‘self’ is mistakenly attacked, we see the development of auto-immune disease, and when ‘non-self’ takes control we fall prey to acute or chronic infections.

Portions of substances the immune system can recognize are called epitopes. T cells recognize epitopes that are bound to Major histocompatibility complex (MHC) receptors. All vertebrate species have an MHC, and therefore the term MHC can be used in reference to any species. Each species also has a unique name for its MHC; in humans the MHC is also called human leukocyte antigens (HLAs) and in mouse it is called H-2.Antigen presentation is mediated by MHC classes I and II. 

Experimental testing of a protein to determine which epitopes bind to a specific MHC molecule requires multiple binding assays; these assays are costly and time consuming. Computational prediction of epitopes can help find or narrow the set of best candidates for vaccine development against a variety of pathogens, including parasites, bacteria, and viruses. Also, computational prediction of epitopes can assist in the development of immunotherapies against a large number of chronic inflammatory diseases and some types of cancer.

However, existing prediction methods are not fully accurate yet. Sometimes they do not always reveal which epitopes are the most naturally immunogenic and thus the most appropriate for inclusion in vaccines and immunotherapies. There are still several unresolved issues in epitope prediction. In particular the prediction of epitopes binding to MHC class II molecules.

My goal is to develop a more accurate method in epitope prediction that could be used as a starting point for epitope discovery.
Home | ResearchCV | Publications | Personal |