In considering isomorphism of unit cells, you also need to consider the accuracy of the cell dimensions determined by indexing programs. Accuracy will be poor for bad diffraction at low resolution from narrow sweeps in low symmetry space groups
Your crystals may be more isomorphous than you think (or less so)
Aimless now does output a matrix of cross-correlations but this doesn’t work well for large numbers of crystals with narrow wedges
Phil
> On 30 Sep 2015, at 12:26, James Foadi <[log in to unmask]> wrote:
>
> Hi Mohamed.
>
> Isomorphism needs to be quantified if we want to classify datasets as "isomorphous". There can be several ways of looking at isomorphism and it is very likely that two researchers will not agree completely on each other's definition.
>
> When using data from multiple crystals with the specific purpose of obtaining electron density maps with good resolution, then it is important that unit cell parameters and atomic positions be as close as possible. In this sense the isomorphism of two crystals could
> be identified by differences in unit cell parameters or by overall RMSD, if the structure has been solved. Prior to structure solution we
> are left only with differences in unit cell parameters as indicators of isomorphism. While all parameters' differences fully account for such measure of isomorphism, it is handy to have a single number to refer to. In BLEND we make use of the so-called "absolute Linear Cell Variation" (or aLCV) which measures the largest difference across related diagonals on the three crystal faces. As the aLCV is given in angstroms this seems to be a sensible way to quantify isomorphism.
>
> There are other factors, though, affecting density-maps quality (anisotropy, map noise, radiation damage, etc). Thus, data from two crystals that are isomorphous in terms of unit cell differences might still not merge well in terms of merging statistics because of these
> other factors. And there are also cases (although I'm not sure how frequent they are) of structures having very close crystal cells but with large conformational changes (like 180 degrees flipping of whole chains). Ultimately the full structure solution highlightes all differences and gives the best answer.
>
> To summarize, and according to my own way of looking at things, when trying to deal with data from multiple crystals:
>
> 1) use data with as close unit cell parameters as possible (measure differences for example using aLCV);
> 2) scale all groups with small aLCV and rank solution according to merging statistics (important in this case are Rpim and CC1/2);
> 3) carry out phasing, model building and refinement for the best solution;
> 4) go back to investigate other solutions starting with the one already obtained as a model (if you have time!)
>
> BLEND is one of the programs with tools to produce such solutions quickly making use of POINTLESS and AIMLESS. It is particularly
> useful when there are several crystals to deal with. In your case (only 4 crystals) you can still use BLEND, (which will include POINTLESS and AIMLESS logs also for the fulll 4 datasets), but can also input your data straight into the CCP4 GUI and obtain the slution for the 4 datasets together straight away.
>
> I hope this helps.
>
> J
>
> Dr James Foadi PhD Membrane Protein Laboratory Diamond Light Source Ltd. Diamond House Harwell Science and Innovation Campus Didcot Oxfordshire OX11 0DE United Kingdom office email: [log in to unmask] alternative email: [log in to unmask] personal web page: http://www.jfoadi.me.uk
>
>
>
> On Wednesday, 30 September 2015, 11:24, Graeme Winter <[log in to unmask]> wrote:
>
>
> Equivalently you can push the whole lot into an aimless scaling run & look carefully at the logs - that’ll probably be more informative than pointless (or BLEND) for relatively large wedges of data
>
> PG622 will help you here it must be said!
>
> Best wishes Graeme
>
>
>
> > On 30 Sep 2015, at 11:10, Tim Gruene <[log in to unmask]> wrote:
> >
> > Dear Mohamed,
> >
> > I would take a look at the plots of pairwise correlation coefficient
> > across resolution. I guess pointless provides these data.
> >
> > Cheers,
> > Tim
> >
> > On 09/30/2015 09:59 AM, Mohamed Noor wrote:
> >> Dear all
> >>
> >> I have a few datasets from multiple crystals with unit cells of about 440, 440, 80 A (90, 90, 120 degrees; point group P 622). While the best 3 clusters together well in BLEND, there is one that does not as its a and b axes are 2 A longer. For the purpose of merging them, can they be considered isomorphic to each other?
> >>
> >> There is an extensive radiation damage (initial resolution from the first 20 degrees is about 3.8 A), so I am hoping to get a good complete dataset.
> >>
> >> Thanks.
> >> Mohamed
> >>
> >
> > --
> > --
> > Paul Scherrer Institut
> > Tim Gruene
> > - persoenlich -
> > OFLC/107
> > CH-5232 Villigen PSI
> > phone: +41 (0)56 310 5754
> >
> > GPG Key ID = A46BEE1A
>
> >
>
>
> --
> This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail.
> Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd.
> Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message.
> Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom
>
>
>
|