Christian Engel wrote:
Dear All,
I am looking for a ccp4 program that reads in cif-files and converts
them into pdb-files, including the CRYST1 card. Can anybody suggest a
solution? I didn't find any in ccp4i, e.g. the coordinate utilities.
I also tried COOT to read in cif-files (downloaded from the pdb
server). For one example it crashed, for others it doesn't colour the
bonds properly if "Bonds (Colour by Atom)" is chosen but shows
all bonds in one colour. Is that a known feature? If I save these
coordinates in pdb-format and try to read it back, I get an ERROR
saying: "Wrong ASCII format of an integer".
Dear wwPDB mmCIF coordinate file users,
I was rather surprised to read that reading a PDB mmCIF file caused Coot
to crash. I was, however, able to reproduce this problem and have made
a fix, which is making its way into binaries as we speak. It was due to
the mmCIF file apparently making reference to chains for which
coordinates (or any other atom or residue attributes) do not exist. Why
is that happening? I don't know, but if it came to a choice between a
broken mmCIF parser and a bogus mmCIF file, I'd plump for the latter.
As to the colours, that is also due, IMHO to a bug^H^H^H issue in the
mmCIF file. The atom names and elements in the mmCIF file are not
quoted so they result in simple left-hand justified strings - such as
"N" - whereas from the corresponding PDB file one would get an atom name
of " N " and an element of " N". These mmCIF simple strings don't match
Coot's (or, i imagine, many other macromolecular file-reading program's)
expectation of atom names and elements - hence it is not recognised and
is represented in grey.
Regards,
Paul.
