Thanks for the suggestion, Tim. :) Someone responded off-list with a similar suggestion. Another suggestion was to download the MTZ from PDB_REDO. These solutions would be great if I only needed one set (or perhaps a handful of sets) of structure factors in MTZ format. Since that's not the case, I should explain the context of my question.
I've been trying to run Phenix.model_vs_data on the whole PDB. That program requires MTZ format, but the PDB supplies structure factors in mmCIF format. I've tried (with multiple versions of Phenix) to use Phenix.cif_as_mtz to perform the conversions, but that's been a less than satisfying experience. It runs rather slowly, and for reasons I haven't yet deciphered, it fails to convert in about 50% of cases. It's not clear to me how much of the problem is rooted in the program vs. my own mistakes. Regardless, I can't afford to lose more time fighting with cif_as_mtz, so I decided to try sf_convert. My understanding is that sf_convert is basically a wrapper for cif2mtz when MTZ is selected as the target format. In theory it's a very convenient wrapper, since it can take PDB model files as additional input to find cell and symmetry information that may be lacking in mmCIF structure factor files. I'm sure I could write a wrapper script of my own, but I was trying to avoid reinventing the wheel. Eric On Wed, Mar 26, 2014 at 5:51 AM, Tim Gruene <t...@shelx.uni-ac.gwdg.de> wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Dear Eric, > > I usually use cif2mtz from CCP4 to import structure factor files from > the PDB into mtz. Works excellently. > > I assume that for some reason sf_convert expects your CCP4 setup file > to be the csh version but finds the POSIX sh version, hence the error > messages. On my system I get the error message '/bin/csh: not found'. > > csh is [flame] deprecated and [/flame] not following the POSIX > standard, so the old scripts cause trouble on modern systems which are > mostly based on bash, or ksh, or zsh, or any other of these variants, > all of are sh-compatible. The best solution to your problem is to use > cif2mtz, and the best answer to your question would be the author of > sf_convert to modernise the program. > > Best, > Tim > > On 03/26/2014 05:00 AM, Eric Williams wrote: > > Does anyone here have experience running sf_convert? I know it's > > not part of CCP4, but it does rely on CCP4 for converting structure > > factors to MTZ format. > > > > When I try to run it, I get errors that suggest that my shell isn't > > bash, but it really is. The program output is as follows. > > > > sf_convert -o MTZ -sf 2hr0-sf.cif -pdb 2hr0.pdb -out 2hr0-sf.mtz > > > > Input file FORMAT=MMCIF Data type is Amplitude (F). If the guessed > > input format (MMCIF) is wrong, please use command: sf_convert -i > > input_format -o output_format -sf data_file > > > > Input File Name = 2hr0-sf.cif : (format=MMCIF) Doing mmcif to mmcif > > conversion.. Number of data blocks = 1 Data_block_id=r2hr0sf, > > pdbid=2hr0 Converting to mtz .. export: Command not found. export: > > Command not found. export: Command not found. export: Command not > > found. ccp4_first_in_path=1: Command not found. export: Command not > > found. export: Command not found. export: Command not found. > > export: Command not found. if: Expression Syntax. > > > > Output File Name = 2hr0-sf.mtz : (MTZ format) > > > > No output file is produced. I recognize some of the output from > > ccp4-setup.sh, but I don't know what's going wrong. Ideas? > > Suggestions? > > > > Eric > > > > - -- > - -- > Dr Tim Gruene > Institut fuer anorganische Chemie > Tammannstr. 4 > D-37077 Goettingen > > GPG Key ID = A46BEE1A > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.12 (GNU/Linux) > Comment: Using GnuPG with Icedove - http://www.enigmail.net/ > > iD8DBQFTMqMhUxlJ7aRr7hoRAifWAJ9c9pn4BRy7XcMfgPRzjdw1mJ69vgCgrF9f > cnNfRbP5XmA4NdUGxXW4S5A= > =HRtQ > -----END PGP SIGNATURE----- >
