Journal
NUCLEIC ACIDS RESEARCH
Volume 42, Issue W1, Pages W59-W63Publisher
OXFORD UNIV PRESS
DOI: 10.1093/nar/gku395
Keywords
-
Categories
Funding
- Ministry of Science and Technology China [2010CB833601, 2012AA020405]
- National Natural Science Foundation of China [31171272, 31200986, 31100956, 61173117]
Ask authors/readers for more resources
Spatial Epitope Prediction server for Protein Antigens (SEPPA) has received lots of feedback since being published in 2009. In this improved version, relative ASA preference of unit patch and consolidated amino acid index were added as further classification parameters in addition to unit-triangle propensity and clustering coefficient which were previously reported. Then logistic regression model was adopted instead of the previous simple additive one. Most importantly, subcellular localization of protein antigen and species of immune host were fully taken account to improve prediction. The result shows that AUC of 0.745 (5-fold cross-validation) is almost the baseline performance with no differentiation like all the other tools. Specifying subcellular localization of protein antigen and species of immune host will generally push the AUC up. Secretory protein immunized to mouse can push AUC to 0.823. In this version, the false positive rate has been largely decreased as well. As the first method which has considered the subcellular localization of protein antigen and species of immune host, SEPPA 2.0 shows obvious advantages over the other popular servers like SEPPA, PEPITO, DiscoTope-2, B-pred, Bpredictor and Epitopia in supporting more specific biological needs.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available