Version history

2.0 The current server. New in this version:
  • Deep learning: TargetP 2.0 is based on convolutional and recurrent (LSTM) neural networks with a multi-attention layer. The deep recurrent neural network architecture is better suited to recognizing sequence motifs of varying length, such as signal or transit peptides, than traditional feed-forward neural networks (as used in TargetP 1).
  • Thylakoid lumen proteins: TargetP 2.0 is now able to predict thylakoid luminal transit peptides (luTPs), which are composed of a chloroplast transit peptide (cTP) followed by a second peptide similar to a bacterial signal peptide.
1.1 The original server. Based on feed-forward neural networks.