Hello Michael and others, Sorry that my statement did cause some confusion. There is nothing wrong with the carve option as such. I also use it very regularly. It is only the way of using which matters.
If you have good density and carve at a sufficient distance around your region of interest, there is nothing wrong. However, if people try to soak-in or cocrystallize a protein with a ligand and get a very noisy map, things get tricky. They might conclude that no ligand has bound, or may try to fit the ligand in some noise or in density of of some partially disordered water molecules. After some refinement, some (biased) weak density may appear and by contouring at a low level and carving tightly, one can end up with convincing looking density. I just did the test and a carving radius of 1.6Å did produce nice looking density from a very noisy map. So my advice would be: -do not carve too tightly so that noise/artifacts in the electron density map near the region of interest are not cut out of the picture. I usually use 2.5-3.0 Å as a carving radius and think 1.6Å is too tight. -contour at a reasonable contour level (not less than 1 sigma on a 2mFo-dFc map). I hope this clarifies things a bit. Best, Herman -----Original Message----- From: [email protected] [mailto:[email protected]] Sent: Saturday, January 15, 2011 2:42 AM To: Schreuder, Herman R&D/DE Cc: [email protected] Subject: Re: [ccp4bb] how to generate density map of selected residues Hello Herman and others, I am somewhat concerned after reading the following comment: "In pymol there is the infamous carve option, which can be used to select density for specific fragments and which can make bad density look good if you carve very tightly around your fragment." I have been using PyMol to generate figures of electron density by creating .ccp4 format maps and displaying them in PyMol, using the option carve=1.6 as recommended by the PyMol Wiki. After reading Herman's reply, I am concerned that the figures I have been making don't accurately represent my density map. I haven't really found any information about what the "carve" option is actually doing. Can anyone comment on the proper way to display .ccp4 maps in PyMol so that they accurately depict the map without "making bad density look good" (or good density look bad for that matter)? Alternatively, is there a better method for making figures of density that does not involve PyMol, or eliminates any potential problems related to the carve option as Herman alluded to? Thanks in advance for any advice, Mike Thompson ----- Original Message ----- From: "Herman Schreuder" <[email protected]> To: [email protected] Sent: Friday, January 14, 2011 7:08:48 AM GMT -08:00 US/Canada Pacific Subject: Re: [ccp4bb] how to generate density map of selected residues Hi Jie, What is the purpose of generating density of a selected fragment? In pymol there is the infamous carve option, which can be used to select density for specific fragments and which can make bad density look good if you carve very tightly around your fragment. I usually delete the residues of interest from the pdb and calculate an omit map. The resulting difference (FOFCWT) map will show a relatively unbiased density for the omitted residues. Best, Herman -----Original Message----- From: CCP4 bulletin board [mailto:[email protected]] On Behalf Of jy Sent: Friday, January 14, 2011 2:03 PM To: [email protected] Subject: [ccp4bb] how to generate density map of selected residues Hi All, I tried the "mask map" function in COOT to generate density map of selected fragment. However, the map looked much better than the original 2Fo-Fc map. For example, weak density now became strong density. I am wondering whether the mask map I generated using default is actually a Fc map? is there a parameter setting to make it reflect the true map? Also, is there a way to select non-continuous residues other than a segment? or what is a better program to do the job? Thanks a lot! Jie Yang -- Michael C. Thompson Graduate Student Biochemistry & Molecular Biology Division Department of Chemistry & Biochemistry University of California, Los Angeles [email protected]
