BioSS%20and%20MRI:%20Computers%20and%20Cattle,%20Statistics%20and%20Sheep - PowerPoint PPT Presentation

View by Category
About This Presentation
Title:

BioSS%20and%20MRI:%20Computers%20and%20Cattle,%20Statistics%20and%20Sheep

Description:

BioSS and MRI: Computers and Cattle, Statistics and Sheep. Iain McKendrick ... Jill Sales (Parasitology) Mintu Nath (Breeding and Sustainable Systems) ... – PowerPoint PPT presentation

Number of Views:27
Avg rating:3.0/5.0
Slides: 41
Provided by: bios2
Learn more at: http://www.bioss.ac.uk
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: BioSS%20and%20MRI:%20Computers%20and%20Cattle,%20Statistics%20and%20Sheep


1
BioSS and MRI Computers and Cattle, Statistics
and Sheep Iain McKendrick (BioSS Principal
Consultant, Animal Health and Welfare)
5th December 2007
2
Outline
  • Introduction to BioSS
  • Overview of BioSS consultancy and research at
    Moredun
  • A sample of some of our work E. coli O157

3
BioSS organisation
  • Staff (38 in total) based at, or regularly
    visiting
  • Kings Buildings, Edinburgh
  • the SEERAD Main Research Providers
  • (MRI, SCRI, RRI, MLURI, SAC)

The role of BioSS is to deliver high-quality
consultancy, training and research in statistics,
mathematical modelling and bioinformatics.
4
BioSS organisation
  • BioSS activities fall into three inter-linked
    categories
  • Consultancy
  • in support of the 4 research programmes
  • externally funded consultancy
  • Research
  • within 3 RERAD funded research themes
  • externally funded research
  • Knowledge Transfer
  • user-friendly software
  • training for scientists
  • postgraduate research training
  • institute-led KT

5
RERAD funding
  • 3 of the RERAD research budget is for BioSS to
    support RERADs research programme through
    specialist advice and training, and to provide
    research in statistics and biomathematics
  • To support Programmes 1-4
  • 600K for consultancy
  • 465K for research

6
Consultancy Team, AHW
  • Iain McKendrick (Principal Consultant MRI)
  • Sarah Brocklehurst (SAC/MRI)
  • Ian Nevison (SAC)
  • Jill Sales (MRI)
  • Mintu Nath (MRI/SAC)
  • Fraser Lewis (SAC)
  • Other expertise available within BioSS

7
Consultancy Team, AHW
  • Iain McKendrick (Bacteriology)
  • Sarah Brocklehurst (Virology Animal Welfare)
  • Ian Nevison (Experimental Design)
  • Jill Sales (Parasitology)
  • Mintu Nath (Breeding and Sustainable Systems)
  • Fraser Lewis (Diagnostic Tests)
  • BioSS staff now on-site at MRI for 4½ days per
    week

8
Consultancy Activities
  • Formal consultancy
  • Review of experiments or grant proposals

9
Consultancy Activities
  • Advisory consultancy
  • Provision of on-site clinics and follow-up
    activities

10
Consultancy Activities
  • Collaborative consultancy
  • Identifying improvements for MRPs, finding
    solutions, facilitating knowledge transfer
  • Substantial contributions to individual projects
  • Blends into research

11
Example Dilution Counts
12
Example Real Time PCR
13
Consultancy Research
  • Consultancy
  • Application of existing statistical and
    mathematical methods
  • Research
  • Modification of existing techniques
  • Development of new methodology
  • Overlap between consultancy and research
  • Work to a standard publishable in international
    journals

14
Consultancy Research
  • Research is applied to an Moredun problem
  • identified via consultancy where existing
    methodology is seen as inadequate
  • Research has to be strategic
  • of sufficient importance to be worth the use of
    resources
  • of sufficient generality to be publishable
  • Research underpins consultancy
  • Consultancy links the stats/maths and
    biological/veterinary research areas
  • We aim for all BioSS scientific staff to work
    both as consultants and researchers

15
Other Criteria for Research
  • Research should fit within one of the three BioSS
    research themes
  • Statistical methodology
  • Statistical bioinformatics
  • Process and systems modelling
  • BioSS staff have, or can acquire, the necessary
    expertise
  • Funding is available (from SG, or elsewhere)

16
Research E. coli O157
  • Bacterium found in rumen, intestines and faeces
    of cattle.
  • Negligible effect.

Photo Stuart Naylor (SAC)
17
In Humans
  • High human public heath risk
  • Bloody diarrhoea or haemolytic uraemic syndrome
  • Wishaw outbreak, 1996
  • Pennington Report, 1997
  • South Wales,
  • September 2005.

Photo BBC News
18
Research Programme
In vitro Adherence Experiments and Analysis
Cattle Challenge Experiments and Analysis
Field Study Design and Analysis
Information about O157 and its epidemiology and
hence viability of different control methods
19
Research Programme
MRI BioSS
SAC, MRI BioSS
SAC BioSS
In vitro Adherence Experiments and Analysis
Cattle Challenge Experiments and Analysis
Field Study Design and Analysis
Information about O157 and its epidemiology and
hence viability of different control methods
20
Field Study
  • Cross-sectional study of 12-30 month old beef
    cattle on 952 farms.
  • 7.9 (6.5, 9.6) of these animals are
    shedding.
  • With 95 confidence, at least 20 of groups
    of cattle contain shedding animals.

Photo Alex McKendrick
21
Field Study 2
Map Giles Innocent, Glasgow
22
Experimental Trials
  • Data from 11 calves.
  • Experimentally challenged with huge dose (109
    colony forming units).
  • Faeces sampled regularly post-challenge.
  • Data from post-mortem sampling along
    gastro-intestinal tract.

23
Adhesion Studies
  • Multiple experiments exploring the effects of
    flagella in promoting adherence.

Photos Arvind Mahajan, MRI
24
Research Programme
In vitro Adherence Experiments and Analysis
Cattle Challenge Experiments and Analysis
Field Study Design and Analysis
?
?
?
Information about O157 and its epidemiology and
hence viability of different control methods
25
Research Programme
In vitro Adherence Experiments and Analysis
Cattle Challenge Experiments and Analysis
Field Study Design and Analysis
Between Animal Infection Models
In vivo Infection Models
Information about O157 and its epidemiology and
hence viability of different control methods
26
In vivo Compartmental Model
  • Birth/Death/Migration process.

27
Ingestion Model
  • Doubly stochastic Poisson process
  • Underlying process reflecting contact of animal
    with contaminated faeces.
  • D density, a function of the environmental pool.
  • K aggregation parameter,value of 8 derived from
    literature.

28
Ingestion Process
29
Mean Population Patterns
30
Assessed Strategies
  • Inhibitory probiotics bacteristatic antibiotics
  • Reductions in birth rate.
  • Bactericidal antibiotics probiotics
  • Increases in death rate.
  • Fasting of animals
  • Change in pH leads to higher birth rate in rumen.

31
Effect of Fasting
32
Benefits
  • Evaluation of scenarios which would be difficult
    and expensive to run experimentally.
  • Could avoid use of animals in initial
    experiments.
  • Informs future biological and veterinary research.

33
Problems
  • Excess heterogeneity in faeces compartment.
  • Experimental trials measure shedding
    longitudinally in cattle.
  • Possible problems with the model.
  • Data not particularly suitable for
    parameterisation.

34
Continuum Model
  • Models growth, advection and diffusion through
    system.

35
Model Formulation
  • Initial Condition
  • Left-hand B.C
  • Right-hand B.C

36
Typical Output
37
Conclusions
  • Some evidence for qualitative differences between
    two different strains of O157.
  • Diffusion coefficient particularly variable.
  • Lack of fit
  • Zero counts
  • Curve constrained to stay too high for too long.
  • Consistent with colonisation (zero advection) for
    some bacteria in some animals.

38
Next Steps
  • Adapt models to allow in vivo colonisation.
  • Parameterisation to draw on MRI adhesion
    experiments
  • Use in vivo model parameter estimates in new
    between-animal model, where unknown parameters
    are estimated using survey data.

39
Final Points
  • BioSS consultancy and research activities both
    add value to MRI and SAC research.
  • Our work would lack focus without collaboration
    with MRI and SAC scientists.
  • Enormous synergies accrue from BioSS position
    outside any particular research
    group/organisation/programme.
  • I look forward to continued collaboration, both
    on Programme 2 activities and others (EPIC).

40
Acknowledgements
  • Scottish Government (RERAD)
  • Joanna Wood (BioSS BBSRC)
  • George Gettinby, Douglas Speirs (Strathclyde)
  • Barti Synge, George Gunn, Chris Low and Stuart
    Naylor (SAC RERAD, Defra, Wellcome)
  • David G.E. Smith, Arvind Mahajan (MRI RERAD)
  • The IPRAVE consortium
  • Giles Innocent (Glasgow Wellcome)
About PowerShow.com