I have a few questions regarding the differences in the secondary structure prediction output from your different servers and from standalone predictions.
First my results :
MIQKAHKRADKGFNDIVAQLYEQEFKTQEKAKYEHIRQAKEKALEEQRISEENQRRVEAEKQA seq
CCHHHHHHHHCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCHHHHHHHHHHCCC i-tasser server (S158976)
CCHHHHHHHHCCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHCC lomets server (S43181)
CCHHHHHHHHHCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHC psspred server (T9522)
CCHHHHHHHHHCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHC i-tasser/psspred local (nr-2013)
CHHHHHHHHHHCHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHC i-tasser/psspred local (nr-2010)
1. i-tasser server and lomets server give different secondary structure prediction. Does the LOMETS server still use PSI-Pred? If so, than I guess it explains the difference.
2. i-tasser server and psspred server give different secondary structure prediction. Is the i-tasser server secondary structure prediction really from psspred, or is it based on the threading results?
3. Is there a way to reproduce the secondary structure prediction from i-tasser server with the standalone i-tasser 2.1 (with new nr database, I can reproduce psspred server, which is good).