-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Dear all,
I received three responses to my Email, those two you saw on the BB
and one from Clemens Vonrhein who pointed me at the program
'check_indexing' from the AutoBuster Suite.
I tried both pointless and check_indexing, and both work equally fine.
I did not try Pierre Legrand's solution because I already had two, but
I am convinced it works equally well.
Pointless has the advantage of doing all combinations automatically,
check_indexing has the advantage of the more concise output. Since
check_indexing only relates all input file to the first one, I applied
the following bash-script to get all pairwise combinations, which
repeatedly pops the first mtz-file from the stack until there are only
two left:
#<-------------------------- 8< snip ------------------------------
mtzs=(1.mtz \
2.mtz \
3.mtz
)
while (( ${#mtzs[*]} > 1 )); do
check_indexing -v ${mtzs[*]}
mtzs=(${mtzs[@]:1})
done
#<-----------------------------snap >8 -----------------------------
The list of mtz-files can surely be expanded (I used the script with
14 entries).
In my particular case pointless had another important advantage which
gave me a great crystallographic lesson:
The beta-angles of all these mtz-files (spacegroup C2) vary by less
than 0.1 degree about 90degrees. While check_indexing takes all the
crystallographic reindexing possibilities into account, pointless also
realised that in this particular case, the crystallographic convention
of beta>90 degrees would render some of the data sets incompatible
with others and therefore reindexed them to beta<90degrees. I was
reminded the molecules don't attend lectures in crystallography and
may therefore not obey conventions. In case of a single data set this
is surely no problem, but in this case it was.
Of course XDS would have picked the correct indexing with beta<89
degrees if only I had provided the same REFERENCE_DATA_SET to all data
sets, but I did not here, and I had to see the reindexing operator in
the form [-h -k l] from pointless in order to realise what was going on.
Thanks a lot for all three suggestions
Tim
On 03/14/2014 11:14 AM, Phil Evans wrote:
> If you assigns them to different datasets in Pointless, then
> Aimless will give you the cross-dataset correlations. By default it
> will scale them to together first, but you can skip that if you
> want
>
> It might not scale well to a large number of files (OK up to about
> 10 I guess)
>
> Phil
>
> On 14 Mar 2014, at 09:33, Tim Gruene <[log in to unmask]>
> wrote:
>
> Dear all,
>
> I am looking for a tool that prints (and preferably plots, e.g. as
> postscript) the pairwise anomalous CC vs. resolution for several
> input HKL-files. xprep does this, but it is interactive and
> requires a fair bit of typing. Since I have a large number of
> HKL-files from XDS, I would like to script that and then flip
> through the pages of the postscript-plots.
>
> I looked into pointless but could not find even a table.
>
> Since there recently were some publications one the use of many
> files for phasing, I though such a tool should exist!?
>
> Best, Tim
>
>
- --
- --
Dr Tim Gruene
Institut fuer anorganische Chemie
Tammannstr. 4
D-37077 Goettingen
GPG Key ID = A46BEE1A
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
Comment: Using GnuPG with Icedove - http://www.enigmail.net/
iD8DBQFTMG2dUxlJ7aRr7hoRAiHUAJ9meTS1Fgc/xYQ2mdOoViVuB7Y8pgCdErms
cVnEfiLoM5SJXxbr+srO9PU=
=NybG
-----END PGP SIGNATURE-----
|