SARpred, a neural network based method predicts the real
value of surface acessibility (SA) by using multiple sequence alignment. In this method, two feed forward, back-propagation networks are used. The first sequence-to-structure network is trained with PSI-BLAST generated position specific scoring matrices. Further, the initial predictions from the first network and PSIPRED predicted secondary
structure are used as input to the second structure-to-structure network. The input is a single
letter-code amino acid sequence in free format and output is a real value of
surface accessiblity corresponding to the amino acid sequence.
If you are using this server then please site:
Garg A, Kaur H & Raghava GP. (2005). Real value prediction of solvent accessibility in proteins using multiple sequence alignment and secondary structure information. Proteins. 61: 318-24
Click DATA to obtain the PSSM for different datasets and PDB files for independent dataset.
