Dear Fred,
People have already done this for all PDB entries:
- http://eds.bmc.uu.se/eds/ : maps and many crystallographic stats
- http://www.cmbi.ru.nl/pdb_redo : maps and re-refinement. And yes, the
stats and maps do improve most of the time, unfortunately also for
structures that are not "old" (but to a lesser extent).
Something else:
Many people have shown that their 'new' validation tool would have
detected the fraud. Good for them, many 'old' validation tools also show
that there is something wrong with the KM structures/homology models. The
point is that we have never actively looked for such fraud. Another thing
(and this annoys the bioinformatician in me) is that many validation tools
do not provide validation reports of the entire PDB ready for someone else
(who like these things) to data mine. If we spend so much time discussing
which experimental data should be made available, then surely validators
should give the right example.
Fortunately there are some good exceptions/exmaples (EDS, PROCHECK, and
WHAT_CHECK) and the PDB is doing its best to make more validation
information available.
Cheers,
Robbie
>In fact, if you want the truth, what I now do is to download the sf.cif
>file, use ccp4i to regenerate the MTZ (with the same Free R-factor flags
>that the authors have used for structure refinement). Then feed that into
>Phenix for a few rounds of positional and temperature factor refinement
>with TLS (very few, I am not really re-refining the structure) and look at
>the resulting e.d. maps. If the structure is "old" then I get improved
>model statistics and (I assume) improved electron density maps.
>
>Fred.
|