I have a question regarding output I obtained from the I-Tasser server. Basically I ran two models of sedolisins, using he structure 3EE6 as reference. For some reason, one model contains all hydrogens whereas the other does not. Modeling was performed with identical settings, the only difference being the sequence. The latter, SED_A, concerns an endosedolisin, the fist SED_B, concerns a tripeptidyl protease and they form references for two subfamilies of sedolisins. The obtained models show a very interesting conformational change that we can very well link with the observed functional diversification. Basically, the endosedolisin requires a larger binding cleft, which we would need to show. The problem is that with the SED_B model (with all hydrogens draw in the model) I cannot obtain a reasonable prediction of the binding pocket. Is there anyway I can obtain the more simple model for the SED_B as well? (no terminal hydrogens in aminoacids).