Natl. Cancer Inst. & Argonne Natl. Lab

Transcription

1 CCP4/GMCA Workshop 6/200 Data collec ction strategy Zbigniew Dauter Natl. Cancer Inst. & Argonne Natl. Lab

2 Advance preparations 1. Crystals must be prepared p 2. You must be prepared

3 Advance preparations 1. Crystals must be prepared p 2. You must be prepared because Anything that can go wrong, will go wrong (Murphy, 2000 BC)

4 Work at the beam line - Very hectic and/or very boring - Requires quick respo onses (time is precious) - Full of surprises - technical problem ms usually on Friday evening - your best crystals do not diffract - cryo stream develops an ice block - storm in Connect ticut, no power - Rarely fully satisfying, sometimes rewarding (lesson of opt imism i and perseverance)

5 Data collection process - Involves lots of technical problems - But it is science, not technicality -Easy to screw-up - Pays off to engage your brain - Last truly experim mental step later mostly computing (and writing-up which may be repeated many times - good quality data steps much in many ways make all subsequent easier

6 Beam line selection - Intensity ty and bril liance 2 nd vs. 3 rd generation synchrotron - Collimation, divergence and focusing undulator or bending magnet - Wavelength range - Detector type and size - Crystal characteristics diffraction strength, cell dimensions - Accessibility (APS vs. NSLS) - User friendliness

7 No marvels But synchrotron b eam makes no miracles bad crystal at home will be bad also at the synchrotron This diffraction is so bad how good we did not bring our best crystals (Hamb urg, ug, 1988)

8 Type of experiment Always best to have diffraction data complete, high resolution and accurate but of particula ar importance are: - Native data for refinement highest h resolut tion (multiple l passes) - Molecular replacem ment medium resolution, no overloads - Heavy atom deriva tive medium resolution, accurate - Anomalous (MAD, SAD) modest resolution (radiation damage)

9 Detector and software In general all CCD or IP detectors and all data process sing programs give equally good data (if working properly) Sometimes importan nt is the size of detector front window (e.g. viruses) Some programs are better for particular applications (e.g. d*trek for fine slicing) or more automatic (user friendly) experimenter s expe erience may be more important than data processing program

10 Quality criteria What me eans good data? Quantitatively and Qualitatively Complete Accurate All reflections in the asymmetric or the anomalous unit have to be measured Intensities have to be meaningful and have realistic error estimates (sigmas or uncertainties) Very easy, but not good to collect indices

11 Quantitative completeness of indices Depends entirely on the geometry and mutu al disposition of Reciprocal llattice (crystal) and Ewald sph here (radiation)

12 Ewald construction 3-D illustration of Braggs law: n. = 2.d.sin Ewald sphere represents radiation

13 Asymmetric unit in reciprocal space Asymmetric unit in reciprocal space is always a wedge bound ded by rotation axes (or planes in Laue group): Triclinic Orthorhombic - etc. native anomalous hemisphere - sphere octant - quadrant It is important to know where to start t and how much to rotate the crystal

14 Asymmetric unit in 622 c axis rotation

15 Asymmetric unit in 622 a/b axis rotatio

16 Asymmetric unit in o axial

17 Asymmetric unit in o diagonal

18 Strategy programs Such considerations are easy with crystals in axial orientation, but in arbitrary orientation it may be difficult Good to rely on the strategy programs (This is a minimalist approach, 360 o will always give complete data set, but beware of radiation damage!)

19 Still image

20 Increasing rotationmosaicity in reciproca space max = 180. d -. a

21 Cell length along the beam = 180. d -

22 Overloaded profiles Best, strongest reflections very important fo

23 Overload extrapolated 6 = standard profile fitted on shoulders

24 Missing reflections are never random If missing reflections were spread randomly there would be no serious problem However, for any of the above reasons they are always missing systematically Kevin Cowtan s duck, back-transformed with missing i segment of data:

25 Intensities (and their uncertainties) Very easy to collect indices without intensities however, intensities s should be well measured (accompanied by realistic error estimations)

26 Accuracy criteria of intensities R merge (R sym, R int ) hkl hkl i I i i <I> - I i = I/ (I) - generally > 2, or 50% of reflections > 3 Mulitiplicity generally the higher the better

27 Improved versions of R merge R merge - bad criterion from statistical point of vie (depends on multiplicity) Improved forms (unfortunately rarely used): hkl [n/(n-1)] 1/2 I R meas = i <I> - I i hkl i I (Diederichs & Karplu I i 1)] hkl [1/(n-1)] 1/2 i R p.i.m. = i I <I> - I i I (Weiss & Hilgenfeld,

28 Criteria of anomalous data Friedel mates can be treated as equivalent in scaling but kept sepa arate on output indicators of anomalous signal: - higher h 2 when Friedels Fidl merged -higher R merge when Friedels merged merge w - list of outliers

29 Numerical criteria of anomalous data R anom = + - hkl I - I hkl I + + I - /2 < F> = <F> hkl F + + F - /2 < F> = < F> hkl F + -F - + hkl F hkl ( F) ( )

30 R merge and 2 (Ta 6 Br 12 soaked crysta

31 Anomalous data - outliers

32 Standard uncertainties ( s) 2-D detectors do not measure individual X-ray qua but somethin ng gproportionalp therefore countin g statistics is not valid and s must be corrected for detector gain <I> - I i t-plot = (I) average = 0.0, 0 s.d. = 1 2 criterion - agreement with expectations

33 Multiplicity More measurementss of equivalent reflections lead to more accurat e average and estimation Also scaling and m erging is more effective But beware of radiation damage

34 Radiation damage Typical syndrome of radiation damage

35 Conclusions X-ray data collect tion (with 2D detectors) -scientific process, not technicality - irreversible consequences (often) -even more important due to progress in automation, phasing, refinement etc. Always involves a compromise between time, redundanc d cy, completeness etc. - but it should be a wise compromise