My personal experience is ( I frequently re-refine structures I cite
if all the data for that exist
in PDB) that PDB possesses a significant number of artifacts
unsupported by reality but by the wild imagination only. These
artifacts are originated from the modest, good and excellent
laboratories alike.
They are maybe not as sounding as tracing the protein main-chain in
reverse mode, but sometimes they support quite significant and
sounding conclusions.
I myself suffered frequently on a stage of structure validations by
PDB due to the wrong treatment of the anisotropic
thermal parameters and erroneous Rfactor calculationsdue to that
during structure factors validation. I think this problem is
resolved now or at least I am not bothered by annotators anymore for
that matter.
Personally I dot believe that by fingerposting events of miss-
interpretations and errors that are difficult to catch
will help to resolve the situation. Peer-reviewing of the data that
enter the PDB is unrealistic. Automatic re-refinement of the all PDB
content which is in the tune with modern "high-throughput of
everything" approaches will not solve the problem either. It will
produce a bit better refinement statistics in the best cases. Nothing
can change
human eye interpretation of the electron density until AI problem will
be solved. Responsibility for the correct interpretation of the
structure is of these who publish it and of these who cite it.
To resume I only wonder, why to fingerpost to 3ftt direction, why now,
why in public and why so emotional ?
Dr Felix Frolow
Professor of Structural Biology and Biotechnology
Department of Molecular Microbiology
and Biotechnology
Tel Aviv University 69978, Israel
Acta Crystallographica D, co-editor
e-mail: [log in to unmask]
Tel: ++972 3640 8723
Fax: ++972 3640 9407
Cellular: ++972 547 459 608
|