Hi John

The script runs MTZFIX as a safety check to ensure that the supplied map
coefficients are valid (since in some cases they are not), i.e. calculated
according to the equations in the literature, as shown here:

ftp://ftp.ccp4.ac.uk/ccp4/6.4.0/ccp4-6.4.0/src/edstats/mtzfix.html

So it calculates the acentric Fourier coefficient 2mFo - DFc from the
values of m (FOM), Fo and DFc in the file and checks the result against the
value given in that map coefficient column for every reflection.  Same goes
for the centric Fourier coefficient mFo and the acentric and centric
difference Fourier coefficients.

For this to work the FOM column obviously has to be not only present but
correct as well!  However if we can establish that the map coefficients are
always valid for a given refinement program then it's not necessary to run
this check every time.  Currently there's no way to avoid the script
running this check, and I don't use phenix.refine so I've no idea whether
the coefficients are as per literature or not.

However if you could contact me off-list and send a sample of the MTZ file
(in confidence) it would be a simple job for me to modify the script to add
an option to skip the MTZFIX check so you don't get that warning.
Obviously merely suppressing the warning message doesn't change anything,
so I would like to run some tests on the file to establish that the
coefficients are indeed as per the literature, since if they are not you
will just get garbage scores from EDSTATS.

Cheers

-- Ian


On Fri, 15 Mar 2019 at 18:01, JP Bacik <[email protected]> wrote:

> Dear CCP4bb crystallographers, in regards to the CCP4 program EDSTATS
> using a Phenix refinement output file,  I tried running the file through
> MTZFIX and it gives the error:
>
> ERROR: Labels not found - maybe non-standard labels used?
>
>
> So I tried changing all the column labels to the equivalent in REFMAC, add
> a FOM column, and get the below errors in the mtzfix logfile. Any advice on
> running Phenix files through the command line version of EDSTATS would be
> very much appreciated. I would like to use this to calculate the RSZD+/-
> values for individual atoms.
>
>
> >>>>>> CCP4 library signal mtz:File column type different from type
> expected by program (Error)
>
> raised in ccp4_lrassn <<<<<<
>
>    From ccp4_lrassn: expected type W does not match file type F for
> column FOM
>
>
> C5    0    0   12    8438.9   184.6    1607.0     0.    7559.2     259.6
> 0.896     0.     0.    0
>
> C7    0    0   12    8438.9   184.6    1607.0     0.    7559.2     259.6
> 0.896     0.     0.    0
>
> C8    0    0   12    8438.9   184.6    1607.0     0.    7559.2     259.6
> 0.896     0.     0.    0
>
> C9    0    0   12    8438.9   184.6    1607.0     0.    7559.2     259.6
> 0.896     0.     0.    0
>
> A1    1    1   11    9520.2    88.6    2866.3  -149.     722.9      51.6
> 0.687  -149.  -149.    0
>
> A2    1    1   11    9520.2    88.6    2866.3  -149.     722.9      51.6
> 0.687  -149.  -149.    0
>
>       1    1   11    2866.3     722.9    6541.8   -9494.3
>
> A3    1    1   11    9520.2    88.6    2866.3  -149.     722.9      51.6
> 0.687  -149.  -149.    0
>
> A4    1    1   11    9520.2    88.6    2866.3  -149.     722.9      51.6
> 0.687  -149.  -149.    0
>
>
>
> Passed test C6:  FM = m.Fo
>
> Passed test 10:  FM = FC for unobserved
>
> Passed test 11:  FD = 0 for unobserved
>
>
> FAILED test A1:  FC = 2m.Fo - FM
>
> FAILED test A2:  FC >= 2m.Fo - FM
>
> FAILED test A4:  FD = 2(FM - m.Fo)
>
> FAILED test C7:  FD = m.Fo - FC
>
> FAILED test C9:  FD <= m.Fo - FC
>
>
> FAILED a test on re-take - check data!
>
>
> NOTE: uncorrected REFMAC files fail tests A4 & C6.
>
>
>
> TAKE CARE: Provenance of MTZ file cannot be determined!!!
>
>
> Thanks,
>
> John
>
> ------------------------------
>
> To unsubscribe from the CCP4BB list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1
>

########################################################################

To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1

Reply via email to