Analyzing Your Results
1. Access your images
We recommend using FileZilla Client to access and transfer images as it works across platforms. For first time users, the client should be downloaded. Detailed instructions are available Downloading and using the FileZilla client.
- Host: ftp.hwi.buffalo.edu
- Username: Your user name
- Password: Your password
For Mac OS, DO NOT use ‘connect to server’ option (it will lock you out of the ftp database). In addition to FileZilla, FETCH is also reliable and free for academic and non-profit institutes.
For Windows OS, WinSCP is also reliable.
For Linux there are many options available besides FileZilla.
Images can be accessed for up to three months from the date of your initial read and are stored in an offline form beyond that period. If you have any problems accessing your image files, please let us know so that we can help troubleshoot.
2. Look at your images
For academic, not-for-profit, and government users the Center does not routinely look at your images. The downloaded images can be examined with most image display software. The Crystallization Center has two software options that enable the display of the images:
1. In April, 2021 we released our new software Polo, a python-based Graphical User Interface that implements the MAchine Recognition of Crystallization Outcomes (MARCO) autoscoring algorithm, enables human scoring for images, and links all metadata from the Crystallization Center. MARCO Polo is open-source and on GitHub! Available for download from GitHub.
2. Our ‘classic’ option is MacroscopeJ, which is java-based and available in your ftp account. If you have screened samples at the Crystallization Center in the past, you are probably familiar with this software – it links metadata for each well and enables manual scoring of the experiment.
3. Analyze the data
For most users, it is assumed that you will be familiar with results that indicate conditions that are promising for optimization. For those that are not, and those that have not used second-order non-linear imaging of chiral crystals (SONICC) or UV two-photon emitted fluorescence (UV-TPEF) we strongly recommend becoming familiar with the techniques. A positive (white) signal in the SHG images can be used to verify that the hits are crystalline. Not all crystals will have a positive SHG signal, but if a signal is detected, then the object is most likely a crystal (Haupert LM, and Simpson GJ, Methods, (2011), 55, 379-386). A positive (white) signal in the UV-TPEF images can be used to verify that the hits are protein if tryptophan residues are in the sample. Not all proteins, or protein/cocktail combinations, will produce a positive UV-TPEF signal. The object does not have to be a crystal to have a positive UV-TPEF signal (Madden JT, DeWalt EL, and Simpson GJ, Acta. Cryst. D (2011) 67, 839-846). The addition of these capabilities to visual (brightfield) imaging has been shown numerous times to make the difference between not identifying any potential crystallization conditions and seeing conditions that result in a structure.
Use MacroscopeJ to view and classify your images. Identify any wells with potential crystals (‘hits’) in the brightfield images. Review the cocktail-only image (the control) corresponding to each hit to make sure that the hit is not an artifact that was in the plate prior to the addition of protein. Check the signal in the SHG and UV-TPEF images. We recommend checking the SHG and UV-TPEF images also to identify any ‘hits’ that may have been missed visually. A discussion on different visualization techniques for identifying crystals is available in a paper from the Center entitled “Identifying, studying and making good use of macromolecular crystals“.
The data contains more information than a crystal or no crystal. The phase behavior of macromolecules in solution is well studied and because of the format of the crystallization screening method used there is considerable information on that phase behavior than can guide approaches to recalcitrant samples and the optimization of conditions where crystals have been seen. Detailed advice on this is given in another paper by the Center entitled “What’s in a drop? Correlating observations and outcomes to guide macromolecular crystallization experiments“.
4. Use the data
Select the hits to reproduce or optimize. Reproduce any hits that you wish to verify (A protocol that has been used effectively in hundreds of cases is available here) and study the effect of variables on the optimization. While there are many optimization strategies, one that has been effective in-house has been to vary the concentration of the macromolecule and precipitant as well as the growth temperature in a systematic manner. This is described in a paper by the Center entitled “Efficient optimization of crystallization conditions by manipulation of drop volume ratio and temperature“.
If the work results in a publication we request acknowledgment of the Crystallization Center via the appropriate reference in any methods section. If for some reason a crystal hit is not obtained we are always happy to offer advice.
Analysis and utility software is available:
|MacroscopeJ||Image data can be conveniently viewed with corresponding chemical information using the program MacroScopeJ. The software is java-based to be compatible with most operating systems.
|CheckShow||CheckShow is a multi-platform program written in Java that displays a virtual 1536-well plate annotated by well position or cocktail number with color-coded experiment outcomes. This software is useful for identifying well locations on a 1536 well plate for crystal extraction, or other testing purposes.
|CrossPlate||CrossPlate is Java-based software designed for the analysis of parallel crystallization experiments to test the effects of distinct variables on the outcomes. Examples include studying differences between the crystallization outcomes of control experiments, examining different protein formulations, the comparison of multiple constructs and of complexes versus individual components.
|SlickSpot||SlickSpot.xls is an interactive EXCEL spreadsheet that allows investigators to formulate grid screens near a given phase boundary for a particular detergent as discussed in Koszelak-Rosenblum M, Krol A, Mozumdar N, Wunsch K, Ferin A, Cook E, Veatch CK, Nagel R, Luft JR, Detitta GT, Malkowski MG. (2009) Determination and application of empirically derived detergent phase boundaries to effectively crystallize membrane proteins, Protein Sci. 18, 1828-1839.