Hi Lisa, There are many things you can try and the phaser manual gives a lot of useful information what to do in difficult cases. From my experience, it is quite difficult to find solutions for MR with nucleic acids. I recommend to search only for protein. As a side effect you can use this approach as an "unbiased" quality check. If your solution is correct you expect to see additional density coming up for RNA (assumed that the RNA is ordered). Try different protein models containing either C-term or N-term or both. Homology modeling can help a lot to generate a good search model. Try also to give higher values for rms differences (or lower sequence identity) between search model and target. Do you allow too many clashes as a packing criterion?
Cheers, christian p.s. Of course, check that you work in the right space group. Lisa Wang wrote: > Hi all, > I got one data about 3.0 A, belong to C2 space group. There are two > protein molecules and one 18-nt dsRNA per ASU. The structure of last > 100aa (C-terminal) has been reported, and 400 aa at N-terminalhe has > homology structure with sequence identiy 30%. I try to solve it by MR > with phaser. > I did rotation and tranlation with 18-nt ideal dsRNA first, then search > with C-terminal model. I got about 20 sol after seaching dsRNA with TFZ > 8-10.0. I use all of them to serch the first c-terminal, and second > C-terminal. Ding this process, LLG keep on increasing, and TFZ is > higher than 10. But all have serious clash. It looks like all the sol > are wrong. > > Please give me a hand.
