Compound Management Automation at AstraZeneca

1
Compound Management Automation at AstraZeneca
Ian Yates MipTec 2001
2
Overview

• CM principles, challenges and demands
• Examples of automation used
• 4 approaches taken during the past 5 years
• Astra HTS Centre
• their strengths and weaknesses
• how they matched our basic principles
• lessons learnt
• The future
• new AZ CM demands
• the solution

3
Some Basic CM Principles

• Ensure compound accessibility
• in matrices and individually
• locally and globally
• Maintain compound integrity
• Error free compound handling
• use of automated solutions
• integrated data system
• Efficient compound usage

4
Some CM Challenges

• Intense demand for new drugs
• Increasingly diverse customer base
• Changes in HTS strategy
• New technologies
• Advances in chemistry
• Larger collections and more targets

5
Some AZ CM Demands (1996 - 2001)
gt106
Collection Size
gt105
AZ AZ
20 / month
Number of Targets
2 / month
AZ AZ
6
Example 1 - First Steps

• Zymark system installed 1996
• Rotary arm, 3 carousels, barcode reader, inkjet
  printer, RapidPlate?, reagent dispensers
• Source and destination plates loaded manually
• Used for plate-based compound handling
• Throughput 100-200 (96well) plates/day

7
(No Transcript)
8

• Strengths
• good first system
• accessible, flexible, quick start-up

• Weaknesses
• sample integrity
• no environmental control
• reliability problems
• not integrated with compound storage
• manual load/unload of both source and destination

9
Example 2 - Tailor-made

• RTS Thurnall system installed 1998
• probably the worlds first narrow aisle compound
  store
• Automated plate production unit
• CRS track robot, Tecan Genesis RSP150, inkjet
  printer, barcode readers, lid handlers, plate
  hotel
• Capacity for ca. 9,000 microtitre plates
• Throughputs
• 200 - 300 assay plates (96 and 384) / day
• 500 - 1000 cherry-picked compounds / day
• Environmentally controlled (temp and RH)

10
(No Transcript)
11

• Strengths
• good second system
• better specified, longer lead-time, less
  accessible, less flexible
• reliable
• not the inkjet printer
• integrated storage and processing

• Weaknesses
• throughputs no longer adequate
• sample integrity
• problems still exist despite environmental control

12
Example 3 - Off the Shelf

• Tecan Molbank installed 2000
• Molbank store integrated with Genesis RSP 200
  (Roma Arm, barcode reader, plate hotels)
• Capacity for ca. 2500 microtitre plates
• Environmentally controlled
• temperature and dry N2
• Throughput 500 - 1,000 cherry-picked compounds /
  day

13
(No Transcript)
14

• Strengths
• compact size
• integrated storage and processing
• inert gas environment

• Weaknesses
• reliability problems and relocation have hampered
  use
• samples are locked-in
• sample integrity
• process area not controlled

15
Example 4 - Flexible Friends

• Workstations (Tecan, CyBio, Quadra, etc.)
• library Reformatting
• emergency assay plate generation
• semi-automated cherry-picking

16

• Strengths
• add flexibility to CM operation
• generally reliable
• relatively inexpensive

• Weaknesses
• sample integrity
• no environmental control
• higher risk for errors
• sample tracking
• human error

17
Equipment Use Today
Assay Plates
RTS
Retest Samples
Mother Plates
Retest Plates
Tecan
Molbank
Cherry-Pick Samples
Cherry-Pick Plates
Cherry-Pick Samples
18
Have We Met Our Basic Principles? (i)

• Sample integrity
• environmental control
• non integrated systems
• plate-based
• Accessibility
• in matrices and individually
• large plate production capacity
• inadequate cherry-picking ability
• locally and globally
• compounds are available locally and globally
• many standardisation issues

19
Have We Met Our Basic Principles? (ii)

• Error free compound handling
• use of automated solutions
• automated solutions used
• cant meet the demand
• integrated data system
• integrated system in place
• have been breaks in the chain
• Efficient compound usage
• expert at working at low volumes
• plate-based approach

20
What weve learnt

• Plate-based system
• inflexible, slow and inefficient
• no longer meets compound management demands
• compromises sample integrity
• inefficient use of samples
• Non-integrated systems
• compromise sample integrity
• not complete environmental control
• risk for errors (mechanical and electronic)

21
AstraZeneca CM Demands

• Discovery done on 9 sites
• HTS Centres at 4 sites
• UK, US and Sweden
• Shifts in HTS strategy / demands
• larger collection more screening more
  follow-up
• merging random and rational approaches
• big demand for subset screening
• collection enhancement
• Compound management service required by other
  users

22
The Future

• We have upheld our basic principles...
• but weve learnt a lot along the way
• and there is room for improvement
• We have met AstraZenecas CM demands...
• but these are demands are still increasing
• and the drug discovery process is changing

We need to change our CM process
23
The Ideal CM system

• Automated Compound Management Facility
• very high throughput processing of truly
  individually addressable samples
• accessibility
• all samples individually sealed under inert gas
• sample integrity
• integrated system and individually identifiable
  vessels
• error-free
• individual vessel and low volume liquid handling
• efficient compound usage

24
The Automation

• Total flexibility not feasible (or required)
• not possible to access individual tubes fast
  enough
• Two store strategy
• high speed reformatting store
• high speed replicating store
• Standard Vessel
• individually sealed
• individually identifiable
• individually addressable

25
High Speed Reformatting Store (i)

• Capacity for 2 million samples
• Pick / place gt30k samples / day
• expansion capacity built in
• Outputs
• subsets (1-500,000)
• active follow-up
• working Plates
• Optimised for use with Standard Vessel
• Will be installed at the four AZ HTS Centres

26
High Speed Reformatting Store (ii)

• Design work on-going with RTS Thurnall
• Design based on Compact-Store? and ABB Flexpicker
  ?
• Integrated Reformatters
• small volume dispensing
• septum piercing
• inert gas purging
• sealers, labellers, etc.
• future-proof design

27
High Speed Replicating Store (i)

• Capacity for 16,000 plates
• general screening set at different concentrations
• targeted libraries
• representative sets
• shorter life-time
• Ability to supply 100k samples/day to 5 assays
• Novel plate sealing technology
• Output assay plates
• 384 and 1536

28
High Speed Replicating Store (ii)

• Design work on-going with RTS Thurnall
• Design based on Compact-Store?
• Integrated Replicators
• very small volume dispensing
• sealers, labellers, etc.
• future-proof design

29
Equipment Use Tomorrow
WPS
Assay Plates
Other AZ sites
Compound Solutions
Working Plates
Subsets Follow-up Other cherry-pick needs
30
Conclusions

• Past experience
• the demands on CM groups change rapidly
• all automated systems have strengths and
  weaknesses
• Future
• the AZ drug discovery process needs a secure,
  flexible and global CM system which can
• uphold our CM principles
• accessibility, integrity, error-free, minimised
  waste
• support new screening and other drug discovery
  approaches
• allow implementation of new technologies
• provide a dynamic collection

31
Acknowledgements
32
Questions ??