Please cite our paper: Mishra NK, Chang J, Zhao PX(2014). Prediction of membrane transport proteins and their substrate specificities using primary sequence information. PLoS One. 9(6):e100278..
TrSSP: the Transporter Substrate Specificity Prediction Server   
  About |  TrSSP |  Datasets |  Help
Location:   Datasets

Benchmark Data
 
  • Five-fold cross-validation test dataset: The learning dataset contains 900 transporter proteins classified into substrate specific class study (amino acid transporter, anion transporter, cation transporter, electron transporter, protein/mRNA transporter, sugar transporter and other transporter) and 660 non-transporters according to the annotationinformation available in Swiss-Prot. Both the Swiss-Prot IDs and sequences are provided in FASTA format. None of the proteins has >70% sequence identity to any other substrate class or within same class.
    Click Amino Acid transporter (70 sequences), Anion transporter (60 sequences), Cation transporter (260 sequences), Electron transporter (60 sequences), Protein/mRNA (70 sequences), Sugar transporter (60 sequences), and Other transporter (200 sequences) and non-transporters (600 sequences) to download the training/testing dataset for each substrate specific class.
  • Proteome annotation: We used our best model to predict the transporter at proteome level for Human, Drosophila, Yeast, E. coli, and Arabidopsis. We collect protein sequences from Swissprot, which are full-length and experimentally annotated. Number of protein sequences are 19254, 3198, 21142, 7794 and 12197 respectively for Human, Drosophila, E. Coli., Yeast, and Arabidopsis. Result for Human, Drosophila, Yeast, E. coli, and Arabidopsis are available in table form. "Yes" means belong to that class while "No" means not belongs to that class.
 

 
© 2013-2014 by The Samuel Roberts Noble Foundation, Inc.