Automation of Macromolecular Crystallography at SSRL

1
Automation of Macromolecular Crystallography at
SSRL
The Australian Synchrotron New Zealand Users
Workshop September 2003
Robot still
Aina Cohen, Stanford Synchrotron Radiation
Laboratory, acohen_at_slac.stanford.edu SSRL is
funded by the US Dept. of Energy and the National
Institutes of Health
2
PDB structures May-July 03
HOME SOURCES
SYNCHROTRONS
3
MR v. PHASE MEASUREMENT
MOLECULAR REPLACEMENT
EXPERIMENTAL PHASES
4
EXPERIMENTAL PHASING
MIR
MAD
SAD
AB INITIO
SIR
5
ANOMALOUS SCATTERERS
6
Beamline parameters

• To cover the great majority of samples
• ?

7
Beamline parameters

• To cover the great majority of samples
• Energy range lt6-17 keV

8
Beamline parameters

• To cover the great majority of samples
• Energy range lt6-17 keV
• Fast energy moves

9
Beamline parameters

• To cover the great majority of samples
• Energy range lt6-17 keV
• Fast energy moves
• Resolution 1 eV

10
Beamline parameters

• To cover the great majority of samples
• Energy range lt6-17 keV
• Fast energy moves
• Resolution 1 eV
• Spot size 250 µm - lt50 µm

11
SSRL BL9-2
Good Flux Useful Energy Range (6-16 keV)
Rapid Energy Changes
12
BL9-2 Oversubscribed
13
What Else Do We Have?
14
What Else Do We Have?
15
9-1 11-1
Good flux Access to useful energy
ranges -- 15 minutes to 1/2 hour at
best to change energy
9-1 12500-16500 eV 11-1 10500-15000 eV (9-2
6000-16000 eV)
16
Energy Moves at Side Stations
To change energy at BL9-1 or BL11-1 the following
must be repositioned monochromator theta
table slide (theta) monochromator bend
table vertical
table pitch table
horizontal table yaw
Weight (kg) Q315 detector
140 Positioners 340 Goniometer 80 Robotic
Mounting System 90 Counter Weight 72 Other
Devices 55 Tabletop 225 Total - 1000 kg
17
Energy Tracking Requirements
The mechanical components must be highly
reproducible (better than 50 µm). Most of the
effort to implement this system was in
trouble-shooting and replacing components that
were not to spec.
Reliable Computer Controlled Positioners
To change energy from 12500 eV to 16500 eV, the
experimental table at BL9-1 must move almost a
meter (as measured from the end of the table).
18
Energy Tracking Requirements
Advanced Hardware Control System (DCSS)
BLU-ICE GUI SGI
BLU-ICE GUI SGI
BLU-ICE GUI linux (remote)
Distributed Control System Server (DCSS) Central
Database / Scripting Engine
G a l i l
G a l i l
G a l i l
DHS SGI (fileserver)
DHS NT
DHS VMS
DHS linux
G a l i l
G a l i l
Beam Line Optics
Experimental Hardware
Detector System
Fl. Detector Sensor A/D
19
Creating the DCS script
Table Slide Position (mm) verses Monochromator
Theta
Optimize the beam line at different energies and
record the motor positions
Difference Between Measured and Calculated Table
Slide Positions (microns)
Fit these values to a polynomial function of
monochromator theta.
TableSlide 2052.82
MonochromatorTheta x 165.354
MonochromatorTheta2 x 0.219763
Write a Tcl/Tk script
20
Typical Se Edge Scans
BL9-1
BL9-2
21
The Results
22
Further Automation of MAD Data Collection
Reliable Computer Controlled Hardware
Advanced Control System (DCS)

23
The Scan Tab
24
Automated MAD scans
25
What bottlenecks remain?

• Sample Mounting
• Hutch access is time consuming
• Crystals commonly lost due to human error
• Data often not collected from the best crystal
• Data Collection
• Detector Readout Time
• Exposure Times of 10 seconds or more
• Unreliable Hardware
• Difficult to maintain and trouble-shoot
• Increases alignment time

26
What bottlenecks remain?

• Sample Mounting
• Hutch access is time consuming
• Crystals commonly lost due to human error
• Data often not collected from the best crystal
• Data Collection
• Detector Readout Time
• Exposure Times of 10 seconds or more
• Unreliable Hardware
• Difficult to maintain and trouble-shoot
• Increases alignment time

27
SSRL Crystal Mounting System
28
Cassette Stores 96 Samples
Standard Hampton pins
Mount 3 cassettes at the beam line
ring magnet
NdFeB

Ship 2 cassettes inside a Taylor Wharton or MVE
dry shipper

Store 20 cassettes inside a Taylor Wharton HC35
storage device
29
The Dispensing Dewar
30
The Robot and Gripper Arms
Epson ES553 Robot
Z
U
Vertically opening gripper arms
?1
?2
Cryo-tong Cavity
Fingers to Hold Dumbell Magnet Tool
31
Robot Demonstration
32
Crystal screening tab in BLU-ICE
33
Cassette Tool Kit Supplied
Styrofoam box holds liquid nitrogen for loading
cassettes
(A) Sample Cassette and Hampton pins (B)
Alignment Jig to aid mounting pins into
cassettes (C) Transfer Handle for handling cold
cassettes (D) Magnetic Tool to mount pins in
cassette to test pin size (E) Dewar Canister
replaces stock canister in dry shipping
dewars (F) Styrofoam Spacer keeps single
cassette in place when shipping (G) Teflon Ring
to support the canister in the shipping dewar
34
Feedback
Sensor
ATI Industrial Automation force/torque sensor
35
Automated Calibration
36
Impacts Accelerating Difficult Structures

• Yeast RNA Polymerase II (Roger Kornbergs
  group, Stanford University)

• Transcription of DNA into RNA - key step in gene
  expression underlying all aspects of cellular
  metabolism
• Large 450 kDa complex 10 subunits
• 10 years of data collection refinement of
  crystallization and cryo-cooling conditions
  derivatives
• Regular access to BL9-2 significantly accelerated
  the screening process

P. Cramer, et al. Science, 288, 640 (2000)
37
View of the Robot System on 1-5, 9-1, 9-2 11-1
9-1
11-3
9-2
1-5
11-1
9-1
38
What bottlenecks remain?

• Sample Mounting
• Hutch access is time consuming
• Crystals commonly lost due to human error
• Data often not collected from the best crystal
• Data Collection
• Detector Readout Time
• Exposure Times of 10 seconds or more
• Unreliable Hardware
• Difficult to maintain and trouble-shoot
• Increases alignment time

39
High Speed Detector The ADSC Quantum-315
Installed at BL9-2, BL9-1, BL11-1 coming to
BL11-3

• Fast readout (1 second)
• 10X faster than Quantum-4
• 3 x 3 array of CCD modules
• Large active area (315 mm x 315 mm)
• 50 um pixels in full readout mode
• 100 um pixels in binned mode

40
SPEAR3
The relative intensities of the SMB
crystallography beamlines (1 Å and 0.2 mm
collimation) for the current SPEAR at 100 mA
(measured) and for SPEAR3 at 500 mA (estimated).
41
What bottlenecks remain?

• Sample Mounting
• Hutch access is time consuming
• Crystals commonly lost due to human error
• Data often not collected from the best crystal
• Data Collection
• Detector Readout Time
• Exposure Times of 10 seconds or more
• Unreliable Hardware
• Difficult to maintain and trouble-shoot
• Increases alignment time

42
Unreliable Hardware
43
New Final Beam Conditioning System
44
New Final Beam Conditioning System
75 mm
150 mm
45
Solutions

• Sample Mounting with SSRL Robotic System
• Screen up to 288 crystals without entering
• the experimental hutch
• Feedback systems and calibration checks
• ensure reliable operation
• Many crystals are quickly screened and
• data collected from only the best
• Data Collection Times Reduced
• 1 second readout
• higher intensities
• better focus
• Upgraded Final Beam Conditioning System
• Modular design enables rapid
• replacement of broken components

46
Where do we go from here?
Remote Access

• Automated data collection from the best
  crystals
• Automatic structure solution
• Sample tracking database
• More feedback
• Automated beam line alignment and calibration

47
The Macromolecular Crystallography Group
SSRL Director Keith Hodgson SMB Leader Britt
Hedman MC Leader Mike Soltis
Günter Wolf, Scott McPhillips, Paul Ellis, Aina
Cohen, Jinhu Song, Zepu Zhang, Henry Van dem
Bedem, Ashley Deacon, Amanda Prado, Jessica Chiu,
John Kovarik, Ana Gonzalez, John Mitchell, Panjat
Kanjanarat , Mike Soltis, Hillary Yu, Ron Reyes,
Lisa Dunn, Tim McPhillips, Dan Harrington, Mike
Hollenbeck, Irimpan Mathews, Joseph Chang, Irina
Tsyba, Ken Sharp, Paul Phizackerley
Department of Energy, Office of Basic Energy
Sciences The Structural Molecular Biology
Program is supported by National Institutes of
Health, National Center for Research
Resources,Biomedical Technology Program NIH,
National Institute of General Medical
Sciences and by the Department of Energy,
Office of Biological and Environmental Research.


48
Further reading
For more details of the SSRL mounting robot,
please refer to the published description of the
prototype system Cohen et al. (2002). An
automated system to mount cryo-cooled protein
crystals on a synchrotron beamline, using compact
sample cassettes and a small-scale robot J.
Appl. Cryst., 35, 720-726.   
http//journals.iucr.org/j/issues/2002/06/00/he030
0/he0300.pdf
49
Cassette Tool Kit Demonstration