What happens when a file is loaded?

When a coordinate file is loaded by itself (i.e. just a PDB, no PSF), VMD uses heuristics to replace missing values that would normally be provided by a structure file. If necessary, VMD does a distance-based bond search to determine connectivity. A bond is formed whenever two atoms are within ( $R_1 + R_2) * 0.6$ of each other, where $R_1$ and $R_2$ are the respective radii of candidate atoms. If both structure and coordinate files are loaded, no approximations or guesses are made.

After the molecule is read in, new names are added to the coloring categories, and assigned colors. Next, bond connectivity is established and the molecule is analyzed to identify its components, i.e., to determine which residues are protein, nucleic acids, and waters, etc. A search is then made to connect these into larger fragments of the same type, and summary information is printed to the screen. An example output for BPTI is:

        Info 1) Analyzing structure ...
        Info 1)    Atoms: 898   Bonds: 909
        Info 1)    Backbone bonds: Protein: 231  DNA: 0
        Info 1)    Residues: 58
        Info 1)    Waters: 0
        Info 1)    Segments: 1
        Info 1)    Fragments: 1   Protein: 1   Nucleic: 0

There are several types of fragments. Protein and nucleic fragments are homogeneous; either all proteins, or all nucleic acids. However, it is possible for a protein to be connected to a nucleic acid or some other non-protein. When this occurs, a warning message is printed, as in:

        Warning 1) Unusual bond between residues 1 and 2

These warnings will occur with terminal amino acids, zinc fingers, myristolated residues, and poorly defined structures.