What is the ratio of increase in the unit cell dimensions.

1) The unit cell dimensions doubling in one or more directions means you are losing one or more crystal symmetry axes (into non-crystallographic symmetry).

2) If there is less than 10% change, for a for example 60A unit cell axis this, 6A) this is more likely due to systematic error in unit cell dimension determination at such low resolution (6-7A) and an improvement of accuracy in unit cell dimensions at 4A of your dehydrated crystal. Or like Iva suggests, if you use ions to dehydrate these may fix themselves on the protein and enlarge the frontiers.

What is the ratio of increase in the unit cell dimensions.

1) The unit cell dimensions doubling in one or more directions means you are losing one or more crystal symmetry axes (into non-crystallographic symmetry).

2) If there is less than 10% change, for a for example 60A unit cell axis this, 6A) this is more likely due to systematic error in unit cell dimension determination at such low resolution (6-7A) and an improvement of accuracy in unit cell dimensions at 4A of your dehydrated crystal. Or like Iva suggests, if you use ions to dehydrate these may fix themselves on the protein and enlarge the frontiers.

Depends from percentage of changes, as Julie already said. How you did dehydration? There are different ways.It can be just statistical error, but in some cases "dehydration" doesn't shrink your cell significantly, you will have slight difference from cycle to cycle because of internal rearrangement of residues\loops of molecules in the area of contacts. By the way, if you did only one cycle, after second one you can have smaller cell parameters. You can play with dehydration parameters if you have enough crystals and time.

Thank you (Greg, Iva, Julie and Valeriya) to answer my questions and the discussions. The change of the unit cell is about 5% bigger. It was my first attempt on the dehydration. The way I did dehydration was a modified protocol from a paper: 1) Crystals were transferred to a hanging drop with buffer containing 5% more ppt than the mother liquor. Then leave it soaking for 24 to 48 hours, and keep increasing the mother liquor buffer with two-step increasing by 10% and 15% ppt, leaving the crystals to equilibrate against the reservoirs for another 24hr each step. What is your experience on dehydration? How did you do it? I am going to repeat the treatment and check the unit cell again and hopefully to get better resolution.

What is your precipitant? You already have good result, 4 A instead of 6! Remember, "best is enemy of good" . Did your try annealing? I like combination of dehydration and annealing but it is not for every crystal. Protocol is in Samygina, et al.,(2000) Acta Cryst. D56, 595. Every protein crystal is unique. If you have large protein with a number of flexible parts you can have double effect - statistic error+ rearrangement of flexible loops in molecule. Or even something more complicated.

I think strongly and observed that the increase of volume is due to the decrease of h-bonding and specially to broken bridges always présents With pseudo Free water molecules.

Is it possible that there is a rearrangement of the remaining molecules (for example a change in the conformation)

so the packing is very different from the hydrated compound, and the cell parameters may increase at least in one direction

of the loss of water involves disorder of the remaining molecules ?

Hey!

5% error in one dimension could be quite likely an underestimation of the unit cell at low resolution. This usually comes along with a high mosaicity. Did your dehydration experiment also improve the mosaicity or just the resolution?