Аннотация:Hematopoietic stem cell transplantation (HSCT) is widely used for treatment of haematological malignancies. Both raft versus tumor (GVT) effect and potentially lethal graft versus host disease (GVHD) may develop after transplantation. Current approach to limit GVHD, is based on donor selection and immunosuppression. But as alloreactivity is not restricted to the HLA locus, up to 40% of fully HLA-matched transplantations are complicated by severe forms of GVHD. Non-synonymous single nucleotide polymorphisms (SNPs) in the genome coding regions can give rise to polymorphic peptides – minor histocompatibility antigens (MiHA) – recognized by the immune system as foreign.Limiting alloreactivity to SNPs in hematopoietic genes could allow to discriminate between GVHD and GVT. Thus information of which polymorphisms are immunogenic is of great importance.So far only a limited number of MiHA have been discovered. Methods for verification of SNP immunogenicity are slow and laborious so there is a need for algorithms predicting potentially immunogenic polymorphisms.We have developed TIE predict (TransImmuno Epitope predictor) which is a comprehensive pipeline for in silico prediction of MiHAs based on exome data of donors and patients. This pipeline utilizes several algorithms for peptide to MHC binding prediction, which increases the overall accuracy of prediction. TIE predict is linked to TransImmuno Epitope database holding the data on common genetic poymophisms genomic sequences and annotations. We have tested this approach on publically available genetic data and currently starting to use it on patients and donors undergoing HSCT in combination with immunogenicity assays.