SHELXL: Frequently Asked Questions

General

Q2: How do I transfer Rfree flags from XPLOR to SHELXL ?

Q3: How do I produce data-files containing I's or F's with the same reflections marked for Rfree ?

Q4: After using SHELXL for refinement using isotropic B's, my rms deviation are too high. What can I do ?

Q5: How can I move from program X/Y to SHELXL ?

Q6: Why are my R-values much higher after migrating from program X/Y to SHELXL ?

Q7: How do I tell SHELXL about a Se-Methionine in my sequence ?

Q8: Is SHELXL really the best program to refine my structure against my 1.1 A data ?

Q9: Is there any references I could use to compare to my own refinement ?

Q10: Is SHELXL using standard pdb-format ?

Q11: From the CCP4-BB (Anthony Addlagatta): 'Is there anybody who knows 'how to convert the SHELX map files to view in QUANTA?'

Q12: How do I print R-value versus resolution ?

Q13: SHELXPRO does not want to calculate my map and tells me: "** Memory too small for Fourier calculation **". What can I do ?

Refinement

Q15: I have two molecules in the a.u. each containing a heme group and data to 1.1 A. What shall I do ?

Q16: Should I refine occupancies of water molecules ?

Q17: How do I make sure that the correct scattering factors are used ?

Q18: When I try to refine my protein as a rigid body using AFIX -statements, I get 'NON POSITIVE DEFINITE' messages. What is going on ?

Q19: Can you send us the correct scattering factors for ions (Ca2+ and Mg2+) ?

Q20: When I use OMIT_* $H, I get an error-message: ** CB_306 CANNOT RIDE **. What is wrong ?

Q21: What about the HOPE command ?

Q22: When I include my high resolution data, SHELXL produces nan's in some places and then crashes. What is going on ?

Q23: How much memory do I need for the matrix inversion ?

Q24: How do I set up the matrix inversion job ?

Q25: How do I make sense of the numbers of parameters ?

Q26: What is the number below the atom name after running a least-squares matrix inversion ?

Q27: Some Waters are moving away from the centre of density toward the corner of the density. What is going on ?

Q28: My substrate is pushed away by anti-bumping restraints that do not make sense to me. What is happening ?

Q29: Our data/parameter ratio is about five, can we get rid of some restraints ?

Q30: What should be the final R-values of a protein structure at atomic resolution

Q31: Is it possible to refine a structure on two or more data sets measured at two or more wavelengths?

Q32: SHELXL-93 gave esd's for bond angles involving riding hydrogen atoms, SHELXL-97 doesn't - is this a bug or a feature ?

Q33: How do I include the anomalous signal of Se-atoms into my refinement ?

Q34: I have been warned that SHELXL was not really written to refine at 2.4 A resolution. Can I do it anyway ... ?

Q35: After a few small changes, my refinement jobs against 0.83 A data all over sudden always blow up with an error message ** REFINEMENT UNSTABLE **. How can I fix this ?

Restraints

Q37: I have to do a refinement of a protein containing some unusual cofactors and ligands. What is the best approach to create a parameter file ?

Q38: How do I restrain an SCN molecule to be straight ?

Q39: Any suggestions for restraints for glycerol ?

Q40: How do I straighten out an azide molecule ?

Q41: How do I generate a restraint across a symmetry element ?

Q42: How do I generate restraints for a linear arrangement of atoms, i.e. a cyano-group ?

Hydrogens

Q44: Can I put Hydrogens if I have good data to 1.6 A ?

Q45: Will putting Hydrogens introduce more parameters into my refinement ?

Q46: How do I attach hydrogens to waters ?

Q47: I have 0.95 A data and I can see hydrogens in 2Fo-Fc maps. How do I make hydrogens anisotropic ?

Q48: When I add hydrogens, SHELXL shows a a warning "** BAD AFIX CONNECTIVITY: N_1 BONDS TO CA_1 CD_1 **. I did specify the N terminus when I ran shelxpro in the first round of refinement and I checked that I had the HFIX 33 N_1 statement before all other HFIX instructions. What else should I do?

Q49: The pdb file generated using the 'G' option (from .res) in shelxpro doesn't have a distinction for e.g. the three gamma hydrogens of Ile (they are all named 'HG2'). Bug or Feature ?

Q50: After adding hydrogens, SHELXL complains: "** BAD AFIX CONNECTIVITY: N_1 BONDS TO CA_1 **". What is happening ?

Q51: The hydrogens in my pdb-file have B-values of -94.75 A^2 - what is going wrong here ?

Disorder

Q53: How can I find disordered parts of my protein ?

Q54: How do I get a disordered residue under control ?

Q55: What do I do with a sidechain for which the electron density is a complete mess ?

Q56: The most important atom in my structure is the OG of a Serine. It is not really in one position, but no matter what I do I cannot find any density for a second conformer. I would like to locally release the SIMU restraint to let it the OG loose a bit. But it does not work ...

Q57: I have a sidechain in two alternate positions. A water molecule is consistent with position one, but will clash with position two. What can I do ?

Q58: I have a disulfide that looks like it might only be partially in the reduced state. How can I model it ?

Q59: I have a Lysine in my structure that sometimes forms a Schiff base and sometimes not. How can I model this ?

Q60: I managed to partially photoactivate my crystal, can I somehow model the ground state and the activatated state simultaneously ?

Q61: How can I interface with O ?

Q62: How are alternative conformations marked in the pdb-files written by SHELXL ?

Q63: How can I work with alternative conformations in O ?

Q64: Why can't SHELXL use 0/1, 1/0, 0-999/0-999 to flag reflections belonging to the free/work set ?

Q65: Why does SHELXL not model my low resolution data, even if I use the bulk solvent correction ?

Going Anisotropic

Q67: I have data to 1.35 A for a protein containing a heme. Rwork drops by 3.2 and Rfree by only 1.4 % on going anisotropic (approx 77000 data / 33000 parameters). What shall I do ?

Q68: I have 1.6 A data, Rfree drops by 0.2% on going anisotropic. Is this significant ?

Q69: On going anisotropic Rfree converges after 3 cycles, Rwork only settles after about 10 cycles. Should I worry ?

Q70: Rfree doesn't go down all that much on going anisotropic, but the maps are much better. Can't I continue with anisotropic refinement ?

Q71: I get loads of 'may be split' messages. What does this mean ?

Q72: Some of the Uij's in the .res file and in the .pdb file match, although they are in different order, others are completely different. Is this a bug or a feature ?

Q73: Upon PAVARTI analysis of Shelxl anisotropically refined protein structures, I find the majority of atoms with large anisotropy ( axis ratio < 1:5 ) are from residues with alternate conformations. Could this be a systematic problem due to the refinement of both occupancy and anisotropic thermal parameters together?

Bulk Solvent

Q75: I get very high R-values especially at low resolution (with Fobs systematically weaker than Fcalc). Upon refining the Extinction coefficient in SHELXL, R_work and R_free decrease dramatically. Is this Extinction for real ?

Twinning

Q77: Does SHELXL calculate estimated observed F-values for a single twin component ?

Q78: Where can I read about twinning ?

Ab initio phasing

Various

Q81: Where can I find information about SHELXL ?

Q82: How can I display the pdf-version of the manual ?

Q83: I am trying to understand the number of parameters refined from my number of atoms etc., but it doesn't make sense ... What is the problem ?

Q84: I am following exactly what is said in the manual, but SHELXWAT always stops with "** Unsuitable format for file try7.res **". What can I do ?

Q85: I have good 1.9 A data. Things look fine but Rfree is 10 percent higher than Rwork. What can I do ?

Q86: My systems administrator accidentally killed my shelxl refinement at atomic resolution on the last cycle of a 10 cycle run. What is the proper way to resume this refinement?

Q87: I am having problems refining a complex structure with a ligand near a twofold crystallographic axis. After the refinement the geometry of the ligand is wrong.

Q88: My structure contains 2656 amino acids and I'd like to refine it with SHELXL. When starting, I get an error message: "** TOO MANY ATOMS REFERENCED IN A SINGLE DELU INSTRUCTION **" What is happening ?.

Q89: How do I find places where I can improve my model ?

Q90: What criterion does SHELXL (both 93 and 97) use for defining equivalents as 'inconsistent'?

Q91: When I read an O-map produced by SHELXPRO, O complains: "Couldn't read map file header." Am I doing something wrong ?

SHELXD

Q93: What is the reference for shelxd ?

Analysing Structures

Q95: I have noticed that SHELX calculates a different solvent content for my crystal than is calculated by using the Matthews method (or using CCP4 RWCONTENTS or CNS). What is the story.