“School of Biological”
Back to Papers HomeBack to Papers of School of Biological
| Paper IPM / Biological / 14125 |
|
||||||
| Abstract: | |||||||
|
Identifying of B-cell epitopes from antigen is a challenging task in bioinformatics
and applied in vaccine design and drug development. Recently, several methods
have been presented to predict epitopes. The physicochemical or structural
properties are used by these methods. In this paper, we propose a more appropriate
epitope prediction method, LRC, that is based on a combination of
physicochemical and structural properties. First, we construct a graph from the
surface of antigen, then by using the logistic regression, we model the physicochemical
and structural properties and weight each vertex of the graph. Finally,
we utilize a clustering method, MCL, to cluster the graph. The e�?ectiveness
of the proposed method is benchmarked using several antibody-antigen PDB
complexes. The results of LRC algorithm are compared with other methods
(DiscoTope, SEPPA and Ellipro) in terms of sensitivity, speci city and other
well-known measures. Results indicate that applying the LRC algorithm improves
the precision of prediction epitopes in comparison with mentioned methods.
Our LRC program and supplementary material are freely available from
http://bs.ipm.ir/softwares/LRC/.
Download TeX format |
|||||||
| back to top | |||||||
