Signal Peptide Predictors help page

Signal Peptide Predictors predict Signal sequence parts and Signal cutting sites either for Prokaryotic or for Eukaryotic species, according to the selected method of prediction. A Signal Peptide is a specific N-terminal sequence region with structural and physicochemical properties that permits protein membrane anchoring, insertion and translocation, most likely mediated by recognition mechanisms.
The class has four predictors, three for Prokaryots and only one is Eukaryots. This is due both to their different speed of evolutionary divergence and to the different number of representatives: Prokaryots grow and evolve rapidly so their Domain is wider and also more diverse in terms of species. than that of the Eukaryots. i The three Prokaryotic class are specific for Gram Positive, for Gram Negative and for the Gram Negative Escherichia Coli. Gram Positive and Gram Negative differs widely in their membrane structure and chemical composition, leading to detectable difference in their Signal Peptides. The Escherichia Coli predictor must be considered more a refinement of the Gram Negative class of prediction than a new predictor. This refinement is due to the biological importance of this organism and to the wide knowledge of its protein ensemble.
All of them are Neural Network-based. There are two Neural Networks (NN), one dedicated to the Signal prediction and the other to the Cutting site prediction. The NN for the Signal has a one-state output, representing, for each residue, the likelihood to belong to a Signal.

  • Target: A Signal Peptide detection is considered valid only when both a Signal and a Cutting Site are found. The simple detection of the Signal without its corresponding cutting site, is not sufficient at all and will never bring to an effective predicted Signal Peptide.
  • Output: The Output is presented graphically in two colors, according to the legend at the bottom of the image, one color each for the predicted Signal Peptide and for the Mature Protein (the protein, as it is after the Signal Peptide cutting process). The Signal likelihood is also plotted as a continuous line interpolating each discrete positional value.
In the table, the results for the segments are labelled accordingly to their predicted feature:
  • Signal - for the Signal Peptide
  • Mature - for the rest of the Protein
Just on the top of the first image, the description line could contain additional information indicating the detection of a prokaryotic lipoprotein. This kind of proteins have special cutting sites, and special aminioacidic patterns around them, that permit lipo-anchoring.

For further information see: Nielsen H., Engelbrecht J., Brunak S. and von Heijne G.; Protein Engineering 10, (1997), 1-6.