VMD 1.8.5 Bugs, Limitations, and Workarounds

General

• Some nucleic acid structures are not correctly recognized by VMD and will not be drawn in ribbon or cartoon representations (PDB code 1S72 is such an example) It should be possible to workaround this problem with a script to adjust atom names and the new 'mol reanalyze' command. Update 1/2007: VMD 1.8.6 includes a new structure analyzer which addresses this and some similar problems. An ultimate solution is still in the works, since atom and residue names for RNA structures are quite variable and the old algorithm was sensitive to these names.
• VMD will not draw ribbon or cartoon representations of cyclic proteins Update 3/2007: A test version of VMD is now available which can correctly draw cyclic proteins. Until the new code is fully tested it will remain an experimental feature, but we are happy to provide it on request.
• VMD will not draw ribbon or cartoon representations of polysaccharide and carbohydrate chains. Update 1/2007: We are working on new carbohydrate-specific representations which will become available in a future version of VMD. If you're interested in testing the new carbohydrate rendering code, we can provide it on request.
• An incorrect PDB accession code or download failure causes a 0 atom molecule to be loaded, though error messages are printed to indicate the problem. Delete the extra molecule.
• VMD contains a bug in the old "cartoon" representation, where there are sometimes gaps between tubes and beta sheets. Use the "NewCartoon" representation which doesn't have these problems.
• None of the external renderers support 3-D texturing "color by volume" yet.
• The Tachyon and LibTachyon renderers sometimes render an incorrect image when several user-defined clipping planes are activated simultaneously. Update 3/2007: This problem is fixed in VMD 1.8.6.

Microsoft Windows Specific

• The Win32 versions of VMD do not include Python yet, for reasons explained here, and on the Python web site.
• Some combinations of video drivers and window operations can cause VMD windows not to be redrawn properly when they are uncovered. (believed to be an FLTK bug with expose events generated by the Win32 text console window) We believe this problem will be solved once and for all when we replace the Win32 text console window with our own (based on TkCon, or FLTK).
• The 'vmdmovie' plugin has problems compressing movies into directories with names containing spaces, due to a bug with the "VideoMach" program used to convert movies on Win32.

MacOS X Specific

• The VMD PDB parser will not accept files with lines terminated by carriage return characters (a tendency of Mac text editing programs).
• The Mac version of MSMS appears to misbehave when generating a surface for some molecular structures. Specifically, loading the PDB '1hzx' does not produce a surface on the Mac, though it works fine on other platforms. Note: This is not a VMD bug, but is noted here for reference.
• The MacOS X version of STRIDE fails to compute the secondary structure for the system "1S5L" with the selection "protein".

Unix Specific

• Some SGI systems will return a stereo visual that only works with "above/below" stereo, which is not supported by VMD. In this case, VMD will incorrectly report that stereo is available, though it isn't really available.
• The 64-bit SGI IRIX version of VMD includes a 'psfgen' plugin that crashes when used. This occurs when VMD and the psfgen plugin are compiled with certain versions of the SGI MIPSPro compilers. We are working on a resolution to this problem.