Dear BB, I hope you can help. I have been trying to density modify my potential SAD solutions but I think it would help if I could conceptualize the arrangement of molecules in my crystals and it is making my head hurt.
I have 2 crystal forms. I only obtained a single crystal for the first form which I have never been able to reproduce. It processes well to ~3A in P6122/P6522 (a=52.3, b=52.3, c=216.9; 1 mol/AU) and C2221 (a=52.3, b=90.5, c=216.9; 3 mol/AU), systematic absences are convincing and there does not look like any twinning issues. MR has not worked but the closest model has less than 20% similarity. A self rotation function shows strong peaks along x, y and z (k=180) and weaker ones along z (k=60 and 120). In a slightly altered condition I can obtain reproducible crystals although the space group has now changed. I have collected native and iodide-SAD data to ~3.2A on these crystals which now index in P222. Systematic absences suggest they are P212121 (a=52.61, b=88.261, c=212.072 in the SAD data), however, they contain pseudotranslational NCS (0.5,0.5,0.03) which is ~40% of the origin, so it may be another choice of P222. The self rotation function is similar to the previous data. As these cell dimensions and that of the C2221 data are almost identical it looks like there has been a slight shift in the lattice arrangement causing the pseudotranslational symmetry (therefore 6 mols/AU). The native and SAD data are isomorphous within 1% but I get much better statistics from phasing with SAD alone. The anomalous signal is reasonable to 4.5-5A but I do get some potential solutions using PHENIX (FOMs ~0.5, BAYES-CC ~40, scew 10-20). Whilst I can see solvent/protein boundaries, at this resolution it is hard to tell what is going on and I am finding it hard to trace any sheets (it is an Ig-like fold) and I think some NCS averaging may do a world of good. The problem is how do I apply this with improper NCS - i.e. how is the 6-fold NCS from the self rotation related to the translational NCS. Also am I missing something which is blatantly obvious which can help me improve these maps and hopefully push the resolution out to its full potential. I might be asking for a lot with this data but if you can help it would be much appreciated - and if I have left out any details please ask away. Many thanks James Dr James Garnett Division of Molecular Biosciences, Imperial College London Level 5, Biochemistry Building, South Kensington, LONDON, SW7 2AZ, UK.
