We have recently encountered issues that are troubling us with regard to the self consistency and therefore reilability of the 3D model prediction using i-TASSER. I ran a ~230 amino-acid sequence WT1 and another sequence which was WT1+KHELL6His added at the C-terminus and received models that seem to be quite different, the secondary structure prediction matched reasonably except near the N-terminus. To further check this result another altogether different sequence of ~140 amino acids WT2 and WT2+ xxx6HIS were run by a colleague of mine. In this case the secondary structure prediction were consistent but again the models except one varied a lort. Starngely onlly one out of 10 top templates used matched. I provide below the links. We are extremely puzzled by these results since just adding a few residues at the C-terminus should not produce such, very different results!!!
Thank ypu for your time
WT 1 http://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S202819/
WT 1 + KLELL6HIS http://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S202922/
WT 2 http://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S203132/
WT 2 + ASLEYAGHHHHHH http://zhanglab.ccmb.med.umich.edu/I-TASSER/output/S203170/
Regards
Alok K. Mitra