CHAINS=(A)can be used to select all atoms in a PDB file that have the same chain-letter, here A.  Some PDB files contain more than one protein. If, for example, a PDB file contains two proteins plus associates small molecules, e.g., H2O or [SO4]=, and all atoms relates to the first protein are assigned chain-letter A, and all the others are assigned chain-letter B, then if CHAINS=(A)is present, only those atoms with chain-letter A will be selected. In rare cases two or more chain-letters might be needed, in which case the letters needed would be specified as, e.g., CHAINS=(AB). or CHAINS=(CDEF) Up to four letters can be specified.

A more common use of this keyword would be when there is, or will be, a need to re-label the atoms in the PDB file by using RESIDUES. If keyword CHAINS is already present in the data set, and there will never be a need to re-label the atoms using RESIDUES (this is the most likely case), then, to reduce the number of keywords, it is safe to delete keyword CHAINS.  By default, when a PDB file is read in, all chains are automatically identified, and used in constructing the keyword CHAINS. In normal work, this keyword should not be modified, however if there is a need to modify it, the following definition of the text is provided as a guide:

Protein chains are normally labeled A, then B, then C, etc.  If this sequence is correct, then keyword CHAINS=(ABC) is not needed.  If a different sequence is to be used, put that sequence in place of text in CHAINS=(text).

For example, in PDB entry 2V66, the chains are B, C, D, and E, not A, B, C, D; for this system, the keyword, if used, would be: CHAINS=(BCDE).

See also START_RES(text) to define residue numbers.