Applied Research Of Pattern Recognition Technology In The Screening Of Cross-reactive Antigens Between Ureaplasma Urealyticum And Human Sperm

Recent immunology study shows that cross-reactive antigens exist between Ureaplasma Urealyticum (UU) and human sperms. It is of great significance to determine these cross-reactive antigens for the benefit of discovering the pathological mechanism of the male infertility caused by UU and designing contraception vaccines. The traditional way of determing these cross-reactive antigens is mostly based on labor-intensive immunological experiments. However, the involoved protein purification process remains a great technological difficulty, sometimes due to the scarcity of expressed protein. This is the case in determing the cross-reactive antigens between UU and human sperms. Nowadays, the genome sequcencing of UU has been completed and all the ecoded proteins have been determined or predicted. On the other hand, proteins expressed specifically in the human sperms can also be found in the protein database. So it is possible to develop a computer aided cross-reactive antigen detection method using the protein database and pattern recognition technology. This is the very purpose of this thesis.The sequence and structure basis for antigen cross reaction is that there exists identical or similar B-cell epitopes between the cross-reactive antigens. Based on this important principle, a compter aided detection of cross-reactive antigens and their B-cell epitopes between UU and human sperms is devised. First, all the UU proteins and proteins specifically expressed in the human sperms are obtained from the SWISS-PROT database. Second, the identical or similar subsequences of given length are extracted from these two data sets.