Some researchers are not comfortable asking questions about my blog posts in public - i.e. here on this site. I sort of understand this and will of course address your questions nevertheless. I'd like to do this here within the blog and not divulge identity of these respondent. Thank you for the feedback - you know who you are.
1) For the 'Art Delusion': What would you think when multiple droplets set by the same person's hand or a robot on the same reservoir crystallization condition with same concentration of the protein, one droplet appeared the crystals but not the other droplets; or one droplet appeared big fewer crystals but the other droplets appeared small, showing crystals? Art or Science?:-)
This is called non-reproducibility. This is a familiar problem. I can only speculate about the reasons for this non-reproducibility: maybe there are minute details that are different in one experiment as compared to the other. Since crystallization depends on nucleation -which is a critical phenomenon - you may be dealing with issues related to low number statistics.
Clear vote for science though, although not fully understood or resolved.
2) For opt screen pipetting calculator, have you seen Brent Segelke's Crystool before? Guess the site online has been moved or no longer accessible. However, David Cooper's crystallizaton grid calculator is available online if you would like to take a look at it.
Crystallization Grid Calculator
http://ginsberg.med.virginia.edu/grid.html
Yes - I've used Crystool years ago. Is it still available? If that's the case, please forward/post the link!
Thanks for the Crystallization Grid Calculator! I may want to review this tool here on the Crystallization Hits blog.
3) "Those soft-looking, clear objects on the bottom of the sitting drop wells are not 'pre-crystalline material', they're plastic tray manufacturing artefacts". I thought those stains were made by the robot needles to scratch on wells in the plastic plates, not from the original plates. I could be wrong if you did not use robot to setup those plates but saw the stains from the original plates?
It could very well be that you were seeing damages in the plate caused by robot needles punching into the bottom of the well. In my case the artefacts always appeared in the bottom row towards well H12. They were there in any plate I looked at - including empty ones - and were noticeable independent of the dispensation tools I used.
4) It is interesting to see you mentioned "Risky business or, how much protein sample do you need?". What is the lowest number of crystals you screened (including the data collection) to solve a structure?
I think that must thave been for was Sensory Rhodopsin - I had set up maybe 4 or 6 trays (each with 72 experiments). Although, I wasn't involved in the diffraction and structure determination at all. Although ca. 300 setups is nothing to brag about, the quantity of protein used was rather low, ca. 300 ug. These were early day LCP-type nanovolume experiments.
I like your underlying question though - is there anybody setting up low number crystallization tonly ammonium sulfate and PEG3350 dilution series? Sounds like a good strategy if you're short on crystallization supplies / resources. Anybody out there?