Hello,
I want to predict the 3D structure for my adaptor protein which contains 2 domains (both have multiple homologous PDB structures that I used as templates) and a central "region" between the 2 domains with no good hits (tried PSI-BLAST to no avail). With modeller (Chimera), I got reasonable structures where the 2 domains are but a ridiculously long loop for the central region. I-Tasser, on the other hand, gave a decent looking structure but not exactly what I had hoped for (c-score = -0.73, TM-score = 0.62 and est. RMSD = 9.5 A!). Is there a way to model this region separately with a higher accuracy and then merge them all into one final structure? I eventually want to investigate protein-protein interactions but can't model the protein domain-wise because both domains are required simultaneously (one binds to a kinase and the other recruits effector proteins) and there are multiple tyrosines in the "region" that also play an important role in recruitment of other proteins. I am new to this field and can't code much... so any suggestions regarding different approaches are highly appreciated!
Thanks!
