Experiment Philosophies

An ideal system yields crystals that:

• Grow reproducibly in many drops
• Are chunky, not needles
• Consistently diffract to high resolution (<2.5 A)
• Tolerate high solvent concentrations
• Don't stick to the plate
• Don't grow skin on the drop
• Don't require complicated cryoprotection

Nevertheless, non-ideal crystals are feasible too: it simply makes soaking, harvesting and data collection slower (further streamlining is in progress, prioritised according to users needs).

The best way to address prohibitively non-ideal crystals is by finding different crystal forms, whether by using different crystallization conditions, construct redesign, surface mutations, or whatever other trick works.

See the Tips and Tricks page, and our submission to Methods of Enzymology.

When screening fragments using the XChem platform, the aim is to maximise the effective compound concentration exposed to the crystal in order to detect weak binding fragments.

As not all solvents are compatible with a crystal system (e.g. DMSO/ethylene glycol may damage crystals or bind to sites of interest), we advocate evaluation of which solvent is best tolerated by your crystal system.

We provide many of our libraries in both DMSO and ethylene glycol but libraries in aqueous buffers can be supported by the platform.

Historically, because of the overhead of X-ray fragment screening, crystals were soaked with cocktails of 4-10 compounds. Our process makes it realistic to do singleton soaking (one compound per crystal), allowing a higher effective concentration per soak and removes the necessary steps for deconvoluting hits.

The XChem pipeline has been optimised for singleton soaking (as this is our currently recommended procedure), however, libraries prepared as cocktails are accommodated perfectly well; and we can support on-the-fly cocktails, i.e. multiple compounds dispensed into one crystal drop.