To use HECTAR you have to respect the following constraints: The input data has to be in the FASTA format. It is recommended to use only sequence header in the NCBI accession number format (use only characters A-Z, a-z, 0-9 or "_"). Protein sequences should start with a methionine (M) residue. For the moment at each time only 15 input sequences can be analysed. Please insure that your input sequences do not possess any ambigious amino acid residues like B, Z, Y or X. If these residues are present in the N-terminal part of protein sequence the prediction methods implemented in HECTAR might fail. For signal peptide/anchor recognition it is necessary that you enter sequences of at least 25 residues length. To be able to use the additional HECTAR modules (full HECTAR) the input sequences have to have a length of at least 60 residues. Please note: HECTAR returns a score for the detected category of subcellular targeting. This score can be used to assess the confidence in the classification. It ranges from 0 to 1, although it is NOT a probability. The higher it is, the higher is the reliability of the classification.