Conformational B-Cell epitope prediction
[Examensarbete på avancerad nivå]
There is demand for higher quality therapeutic proteins in our society. Medical drug developing companies strive for lower development cost and shorter development time. This require faster and more reliable ways of testing the therapeutic proteins before releasing them to the public. This project aimed to investigate one part of this problem, the conformational B-cell epitopes which is the interface between an foreign molecule (antigen) and an antibody. It was done by development of two different epitope models, which then was used as a base for creation of two training data sets. These training sets were then used to train machine learning algorithms in order to classify areas on molecule surfaces which are prone to be an epitope. Different problems in this research area is discussed and possible solutions is proposed.
Nyckelord: epitope, conformational epitope, artificial neural network
Det verkar som att din webbläsare saknar möjlighet att visa PDF. Klicka Öppna i nytt fönster för att läsa.
file:////usr/local/tomcat/webapps/chex/local/xsl/output/html/normal/body.xsl<Line 340, Column 84>: XML-22021: (Error) Error parsing external document: 'Förbindelse vägras'.