I'd try first Mosflm, a cup of tea and two images. I'll get to the cup in a 
minute.

Read two images or three which are 90 degrees apart (three incase you have a 
cubic space group).
Then take a sip and start manually removing spots which you believe belong to 
the other lattice, then index and see if you get a decent cell which overlaps 
with most of the reflections. It might help to look at a movie of your dataset, 
so that you can distinguish which lattice points belong to which crystal. As an 
alternative approach I believe Mosflm can use JPG images, so convert your 
dataset to JPG and remove those spots which belong to the other lattice in 
Photoshop. This approach might lead you to a "dataset" and provide you with a 
cell, but I think there are better ways of spending time.

Now XDS, you write it failed, why ? Are you indexing over the whole dataset ? 
Do you have the beam position right ? I'd try to give it the P1 cell it 
suggests and then process the whole dataset even if it will be incomplete. Then 
run pointless and see if you get another suggestion for a space group & cell. 
With that information I would then go back into IDXREF.LP and see if a similar 
cell was picked up by XDS and try it with that.
You will have a lot of reflections which will be rejected (because they belong 
to the second lattice) but you should be able to extract sufficient information 
to be able to work with the data.

Next, your crystals, have you tried seeding them to get them grown individually 
? Decreased protein concentration, larger drop volume, different ratios, all 
other usual things that come to mind ?

Good luck !

Jürgen

-
Jürgen Bosch
Johns Hopkins Bloomberg School of Public Health
Department of Biochemistry & Molecular Biology
Johns Hopkins Malaria Research Institute
615 North Wolfe Street, W8708
Baltimore, MD 21205
Phone: +1-410-614-4742
Lab:      +1-410-614-4894
Fax:      +1-410-955-2926
http://web.mac.com/bosch_lab/<http://web.me.com/bosch_lab/>

On Feb 11, 2011, at 11:06 AM, pravinkumar jagtap wrote:

Hi to all,
I have crystals which grow as joint crystals i.e two crystals are joined to one 
another. I tried to seperate the two and collect dataset but the resolution is 
not good( probably I am harming the crystals during seperation). Also I 
collected a dataset with the joined crystals in the loop and I got resolution 
up to 3.5 A on synchrotron. But when I am trying to index this dataset, the XDS 
is failing . I also tried to use mosflm but it is also giving warning message 
that the unit cell parameters are too inaccurate. And this all seems to be 
because of the two crystals in the loop. Can anyone suggest any strategy which 
I should take for indexing this dataset.

Thanks,
Pravin Kumar

Reply via email to