Transmembrane Strand Predictors help page

Transmembrane Strand Predictors (Porin), class contains two different predictors, one Neural Network (NN) based and the other built upon a Hidden Markov Model (HMM). Both of them make use of "Sequence Profiles".

  • B2TMR (NN and Profile)
  • B2TMR-HMM (HMM and Profile)
  • Target: Porins, maltoporins, outer membrane proteins, ligand-gated protein channels and other transmembrane proteins with a beta-barrel structure.
  • Output: Output is presented graphically in two colors, according to the legend at the bottom of the image, one for the predicted transmembrane beta strands and the other for the outer-inner membrane protein sequence parts (loops).
    The output is represented by two predicted states for each aminoacidic position, one representing the probability to be in a transmembrane beta strands and the other to be into a loop. The difference between these two predicted probability is represented as a continuous line interpolating each discrete positional value.
    In the table of the results each segments is separated from the others and dedicated labels help to classify them according to their predicted feature:
    • Internal - for inside membrane surface loops
    • Transmembrane - for membrane spanning segments (beta-strands)
    • External - for outside membrane surface loops
    This protein membrane orientation derives directly from the method only for the HMM based one. For the Neural Network based predictor is applied a further computation based on rules derived from the loop length.

For more detailed information see: Jacoboni I, Martelli PL, Fariselli P, De Pinto V, Casadio R Protein sci 10: 779-787 (2001), Martelli PL, Fariselli P, Krogh A, and Casadio R Bioinformatics Vol.18 no.90001 (2002).