SEPPA 3.0-enhanced spatial epitope prediction enabling glycoprotein antigens. Other Scholarly Work

Zhou, Chen, Chen, Zikun, Zhang, Lu et al. (2019). SEPPA 3.0-enhanced spatial epitope prediction enabling glycoprotein antigens. . NUCLEIC ACIDS RESEARCH, 47(W1), W388-W394. 10.1093/nar/gkz413

cited authors

  • Zhou, Chen; Chen, Zikun; Zhang, Lu; Yan, Deyu; Mao, Tiantian; Tang, Kailin; Qiu, Tianyi; Cao, Zhiwei



  • B-cell epitope information is critical to immune therapy and vaccine design. Protein epitopes can be significantly affected by glycosylation, while no methods have considered this till now. Based on previous versions of Spatial Epitope Prediction of Protein Antigens (SEPPA), we here present an enhanced tool SEPPA 3.0, enabling glycoprotein antigens. Parameters were updated based on the latest and largest dataset. Then, additional micro-environmental features of glycosylation triangles and glycosylation-related amino acid indexes were added as important classifiers, coupled with final calibration based on neighboring antigenicity. Logistic regression model was retained as SEPPA 2.0. The AUC value of 0.794 was obtained through 10-fold cross-validation on internal validation. Independent testing on general protein antigens resulted in AUC of 0.740 with BA (balanced accuracy) of 0.657 as baseline of SEPPA 3.0. Most importantly, when tested on independent glycoprotein antigens only, SEPPA 3.0 gave an AUC of 0.749 and BA of 0.665, leading the top performance among peers. As the first server enabling accurate epitope prediction for glycoproteins, SEPPA 3.0 shows significant advantages over popular peers on both general protein and glycoprotein antigens. It can be accessed at or at Batch query is supported.

publication date

  • July 1, 2019

published in


  • Algorithms
  • Antigens
  • Area Under Curve
  • B-Lymphocytes
  • Databases, Protein
  • Datasets as Topic
  • Epitope Mapping
  • Epitopes, B-Lymphocyte
  • Glycoproteins
  • Glycosylation
  • HIV Envelope Protein gp120
  • Humans
  • Internet
  • Logistic Models
  • Protein Conformation, alpha-Helical
  • Protein Conformation, beta-Strand
  • Protein Interaction Domains and Motifs
  • Protein Processing, Post-Translational
  • Software

Digital Object Identifier (DOI)


  • Print

start page

  • W388

end page

  • W394


  • 47


  • W1