Welcome to NETASA:
Prediction of Solvent Accessibility using Neural Networks


This program predicts the solvent accessibility of amino acid residues in a protein using our neural network algorithm. Neural network was trained using 30 proteins and results were tested on 185 proteins. Detailed results of prediction accuracy on these proteins are available Here.
Brief description is provided in the help document.

Reference: Shandar Ahmad and M. Michael Gromiha, Bioinformatics , 18 (2002) 819-824
To predict ASA states for a protein sequence, choose a state threshold for buried and exposed categories. Three state predictions have two thresholds for buried, partially buried and exposed residues.
0% (Two State)
5% (Two State)
10% (Two state)
25% (Two state)
50% (Two state)
10%,20% (Three state)
25%,50% (Three state)

Please enter your sequence in a single letter code.