Hello,

Version 2.0.6 of the Format Converter and 2.0.7 of Analysis are now live. 
See comments in original email (way) below about what to do if you are on 
Analysis 2.0.6.

One thing to note.  Several people have noted various problems with Link 
Sequential Spin Systems.  Tim put a fix in which will probably resolve 
some of these issues, but there still might be others.

Wayne

Analysis 2.0.6 to 2.0.7 changes:

Added structure validation option to Structure menu. Calls CING remotely 
via the iCing server in Nijmegen.

Added trial version of the Entry Completion Interface for preparing 
BMRB/PDB depositions. We will give details how to avtivate upon request.

Increased the number of types of spectra that you can create synthetic 
peak lists for. This includes HNCO, HNCA, HNCACB, HNHA, HN_H.TOCSY, 
HN_H.NOESY. The rule is that the there must be a direct transfer between 
the measured experimental dimensions. This excludes for example HNcoCA 
where transfers via the CO is not represented by an experimental 
dimension.

Added "Propagate Merit" option to the peak tables. This allows easy en 
masse setting of a peak worthiness. Good for resurrecting disgruntled ARIA 
peaks.

Added calculation and display of peak line widths via peak tables (but 
still doesn't deal with overlap too well).

In Restraints & Violations, added funtion to split a restraint list in two 
according to whether restraints are violated or not violated.

Added "Process All" option to Link NOEs, to assign a whole peak list. Use 
with caution.

Improved file selector, with more space, file history, navigation buttons 
etc.

Added new check to Quality Reports: Warn if only one prochiral of an atom 
pair is assigned to a peak dimension if the other prochiral is very close 
in chemical shift.

Synthetic peaks are now down-weighted (0.01) in the chemical shift 
calculation.

Added automatic tidy of unused dataLocations when spectra are deleted.

Added button in Edit Structures to remove residue validation parameters. 
(e.g. RMSDs).

Added ability to focus on selected residue in Edit Structures.

Structures are no longer selected by default in the Restraints tab. (This 
could be very slow). Updating of the restraints table now made faster, 
including when calculating structural values is enables.

Improvements to the Nexus automatic sequence assignment: It copes better 
with bad data, and the scores in the graph are now unlikely to give 
spuriously high scores.

The scores are now based on both rate of convergence and goodness of 
sequential matches.

Display or spin system strips now takes an optional shift list setting.

Added isotope labelling hilighting to the Assignment Graph.

Added setting of minimum isotope fraction in Edit Assignment popup (only 
used with a labelling scheme).

Generating peaks from Restraints improved: Was missing some hetero atom 
assignments before.

Can now generate restraints for carbon-carbon through-space transfer 
experiments.

Isotope labelling filtering of assignment options now works for resonances 
that have a residue type and atom type but no sequence number.

Added "frequency" and "pulsing frequency" to NMR series condition types.
Added some functionality for 1D shapes.

Make importing Bruker spectra easier by allowing selection of proc files 
several directories up.

Various bug fixes dealing with sampled dimensions in spectra.

On Mon, 6 Apr 2009, Wayne Boucher wrote:

> Hello,
>
> This is just a warning that we are putting out a new release of v2 Format 
> Converter and Analysis pretty soon, likely tomorrow the way things are going.
>
> But for Analysis users there are a few housekeeping notes to get through 
> first.
>
> So the first note is that normally you can do an update from inside Analysis 
> and it will notice that you have an out-of-date version and try to download 
> the latest version.  Unfortunately testing stuff at the weekend I discovered 
> that a bug was introduced into the update code in December which means that 
> in 2.0.6 Analysis this update feature is broken (it will just not get around 
> to asking about downloading the new version). So I've just now put a fix on 
> the update server for 2.0.6 users.
>
> For people who do *not* get this fix before the next release is out (and once 
> the next release is out you won't be able to get it), here is how to fix it 
> in your 2.0.6 code.  In the file 
> ccpnmr2.0/python/ccpnmr/update/UpdatePopup.py on line 77 you have:
>
>    self.update()
>
> Change that to:
>
>    self.updateFiles()
>
> And hopefully you can then upgrade from inside Analysis (well, obviously do 
> not start Analysis up to do the upgrade until after that line is fixed).
>
> Which brings me onto the next point of bureaucracy.  Many people have 
> pre-compiled releases of 2.0.6 Analysis.  Unfortunately in 2.0.6, the update 
> mechanism for you would still download the source code version of 2.0.7 and 
> try and compile that (with the same Python/Tcl/Tk as was used for 2.0.6, at 
> least).  Now this compilation will almost certainly fail for some people for 
> various reasons (e.g. some system header files might be missing, we've 
> already hit that one in another context).
>
> In 2.0.7 I've changed things so that from now on pre-compiled releases will 
> try and download pre-compiled releases.  But for now, anybody who has a 2.0.6 
> pre-compiled release I would recommend that you get the 2.0.7 pre-compiled 
> release yourself from:
>
>  http://www.ccpn.ac.uk/ccpn/software/downloads-v2/
>
> (when it is ready, it's not yet!), instead of using the Analysis update 
> mechanism.  Just unpack and run "python configRelease.py" again (as last 
> time).  But first move the 2.0.6 release because both 2.0.6 and 2.0.7 will 
> unpack into the same place (well, if you unpack in the same directory).
>
> Wayne
>