Not every crystallization trial/work unit produces a crystal. The success rate is under 1%. We are interested in all crystals. Most crystals grown in these trials are not X-ray-diffraction-quality, but indicate the cocktail is close to an optimum chemical condition.

As for what is diffraction-quality, visual inspection alone is not sufficient to tell. Some crystals that look dirty turn out to diffract well, and some gorgeous crystals diffract poorly, or turn out to be salt rather than protein. But large, clear, geometrically-ideal crystals are what a crystallographer will try first.

Christian

Every protein in the study is initially tested against 1,536 chemical conditions. Some proteins will not crystallize at all, some will crystallize under one or two conditions, and a lucky few may crystallize under many conditions. Two different proteins will respond differently to the same chemical conditions, and the optimization process for each will differ. Our lab does deal with image data from optimization experiments, but that data is not being fed to the World Community Grid at the moment. The WCGrid is tackling the hard part, finding crystals in that initial 1,536 screen.

Predicting what cocktail will work on any given protein (prior to any crystallization screens) is still next to impossible. One very exciting benefit of this project is that the data gathered from these millions of experiments, and the patterns discovered therein, will help create a practical science of crystallization.

Adding some human power to the project is a great idea, and one that we are considering for the future.

Thank you all for your contribution to the project!

Christian

Borderline, but X0000081931434200702141932 may have lots of little crystals. Hard to see but the dots look a little more regular and hard edged that the usual crud.

http://www.flickr.com/photos/wombat_56/4731754834/

Christian, I believe I have another image with the micro crystals you mentioned. It is hard for me to tell because it looks like there is a film on the surface. Are the Help Conquer Cancer scientists interested in enough in images like this that we should post them when we see them?

Work unit = X0000080740819200701021144
http://i47.tinypic.com/vg3nt2.jpg

Thanks,
Adam

Found another.

X0000030410523200404151932

Adam