Atom Specification in Chimera versus Midas

While atom specification in Chimera is very similar to that in Midas/MidasPlus, there are some significant differences, which are summarized below.

The differences in atom specification mainly reflect additions that do not affect the process of adapting a Midas script for use in Chimera. However, note that in Chimera, chain IDs are specified differently and mainchain is not accepted as a specifier. A few Midas/MidasPlus commands are unsupported by Chimera.

• The three major levels in the hierarchy (# for model, : for residue, and @ for atom) can each have sub-specifications indicated after "." (submodel, chain ID, and alternate location indicator, respectively). This is a major improvement, because:
  • When multiple MODEL records appear in a single PDB file, each MODEL can now be specified independently as a submodel (for example, #0.2 is the second submodel in model 0). In Midas, the only way to see one structure at a time is to edit the input file.
  • One can specify an entire chain without knowing the starting and ending residue numbers (for example, :.a specificies all of chain A).
  • Atoms in alternate locations can be controlled independently.
• Atoms/bonds that are selected in the graphics window can be designated with the word selected, sel, or picked.
• The name of a picked atom is not echoed in the command line (as in the most recent version of Midas), but entering a plus sign (+) indicates that the plus sign should be replaced by the specification string of the next picked atom (similar to an older version of Midas).
• Selections previously named and saved with Select... Name Selection or the command namesel can be specified using sel=selection_name or just selection_name
• One can use names of entries in the following parts of the Chimera Select menu:
  • Chemistry... element and submenus
  • Chemistry... functional group and submenus
  • Chemistry... IDATM type
  • Residue... amino acid category
  • Structure (which includes the automatic categories for surface calculation: main, ligand, solvent, and ions) and submenus
  The same pattern of capitalization and spaces, if any, as shown in the Select menu should be used. Where there is ambiguity, the parent menu should be included in the specification, for example, "IDATM type.H" or "element.H" instead of "H" alone.
  One can also use the names of custom amino acid categories defined with ResProp and custom categories for surface calculation defined with the command msms cat (surfcat).
• The word mainchain is not accepted as an atom specification; the MainChain tool provides an effect similar to that of chain mainchain in Midas.
• There are several differences in property (attribute) descriptor syntax:
  • In Chimera, property descriptors are termed attributes and are defined at the atom, residue, and model levels. Thus, the attribute indicator slash ("/") must be preceded by the proper symbol (#, :, or @) depending on what level is being described. The symbol does not have to precede the slash immediately (for example, @ca/color=red indicates red atoms named CA). In Midas, all the properties are atomic, so the symbol before the slash is not required unless an atom name is also being used. There are more attributes in Chimera than in Midas.
  • When multiple attributes at the same level are combined in one statement, they must be separated with and and/or or to indicate intersection or union, respectively. In Midas, commas indicate intersection (and for colors only, union) and multiple occurrences of slash indicate union.
  • Not is indicated for yes/no properties with !descriptor_name, and for multivalued properties with descriptor_name!=value. In Midas, ! always precedes the descriptor_name.
  • The operators ~ and !~ can be used instead of = and !=, respectively, to indicate that the subsequent string should be treated as a regular expression.
  • The descriptor_name is case-sensitive; the descriptor value, if any, is case-sensitive if specified with ==, but not if specified with =.