ModelRepository

Joint project of EFSA, Federal Institute For Risk Assessment, DTU and ANSES to create an online model repository.

content

currently implemented models

Nauta EFSA opinion Risk Reduction model of campylobacter in ready-to-eat chicken salad

containing

• 8 Consumer phase models
• 3 Dose Response models
• 1 Relative Risk Reduction Model based on above mentioned CPMs and DRMs
• 3 examples of joined CPM/DRM models

  Status

  done

  Reference

  https://efsa.onlinelibrary.wiley.com/doi/full/10.2903/j.efsa.2020.6090

Lindqvist EFSA opinion

containing qmra model on listeria in several food types

Reference

doi: 10.2903/j.efsa.2020.6092

Status

done, waiting for last check with author

Swart Taenia model

containing Taenia solium risk analysis model

Reference

...

Status

pending: waiting for expert review (Swart, Ganas?)

Ranta models

containing

• chemical dose response model
• bayesian consumer phase model on campylobacter in chicken

  Status

  pending: waiting for expert review (Ranta, Ganas?)

  Reference

  ....

gropin modelling database

containing 997 models in total, which are divided into (old version, needs update!)

• 495 growth models (ALL transferred!)
  • 1 variable - 291 models
  • 2 variable - 99 models
  • 3 variable - 57 models
  • 4 variable - 40 models
  • 5 variable - 2 models
  • 6 variable - 4 models
  • 9 variable - 2 models
• 72 inactivation models (inverse growth models, should be equal to growth model, check first!)
• 131 growth/no-growth models (polynomial fitting necessary?)
• 284 lethality models
• 15 gamma models with interaction (probably not transferrable, check first!)

  Reference

  https://www.aua.gr/psomas/gropin/

  Status

  ongoing transfer, growth models completely transferred (roughly half of gropin DB)

TODO

Gropin

• non functional models
  • sqrt of negative numbers -> ignored by excel, error in R - excluded models
  • zero in denominator if Variable X - Coefficient Y = 0, happens on boundary of range of variable X, place holder solution: range of variable shortened by relative 1/1000
• look into inactivation models (probably similar to growth models, but need to be checked)
• look into GNG models
• look into lethality models

other models than gropin

P#ossas

• -> check up on Cristina
• transfer Possas model(Cristina) to fskx
• discussed to be done in 9 steps
• sending Cristina R scripts for development, afterwards transporting into fskx

DONE 2021

June

Gropin

• 3 and more parameter models
  • 3 variables testing - DONE
  • multiple-variable model transfer code - DONE
• rewrite code: coefficients were introduced as constants -> remove from meta data schema -> hard code into model script
• create visualisation code for all multiple variable models, so each hyper plane may be viewed - DONE
• rewrite visualisation code to accomodate for Matthias wishes to create a more simple visualisation code
  • just a generic visualisation combi plot of 3 pairs of variables-DONE
  • create generic visualisation code for primary model code - DONE
  • rewrite code for new visualisation code to accomodate for all possible hyper plane permutations:
  • 1 plane for views for 2 variables (only 1 possible pair)- DONE
  • 3 planes of views for 3 variables (3 possible pairs) - DONE
  • 6 planes of views for 4 variables (6 possible pairs) - DONE
  • 10 planes of views for 5 variables (10 possible pairs) - DONE
  • 15 planes of views for 6 variables (15 possible pairs) - DONE
  • 21 planes of views for 7 variables (21 possible pairs) - DONE
  • 28 planes of views for 8 variables (28 possible pairs) - DONE
  • 36 planes of views for 9 variables (36 possible pairs) - DONE
  • rewrite code for all possible new simulation settings - DONE
  • permutation formula: number of possible pairs = \fraction{\factorial{n}}{2*\factorial{n-2}} (n = number of Variables)
• create additional simulation parameter set: point estimates of Par - DONE
  • edit sim.sedml and add all new simulation (dynamical depending on number of variables)
  • remove fixated solution - DONE
  • remove _start,_end solution of fixated variables -> just the array - DONE
• rewrite visualisation code to accomodate for users choice of simulation settings - DONE
• default resolution dependent nrOfVariables ceil(nrOfVariables_throot(max(21*21))) (R package pracma::nth_root) - DONE, see in gropin2R: lenOfVarVec
• comment readme about Rechenzeit with full resolution - DONE
• add generic visualisation script as comment for data from secondary model - DONE
  • rewrite generic visualisation to accomodate for more complicated version of visualisation choice (see above) - DONE
• joiner workflow als beispiel (Petra?) - DONE
• check for sqrt/ln/lin mumax as is? - DONE (secondary model wont touch mumax, but retain information about which mode was used in header of resultTotal -> primary model will read this information and recalculate mumax accordingly)
• units:
  • create database for units of variables: T in C, CO2 in g, a.s.o. - created, but not filled with correct information, SUPPORT NEEDED!
  • meta data needs to be adapted for each microorganism separately - OPEN QUESTION!

May

• new model version of gropin -> major adaptations were necessary
• R4.X handles xls sheets differently thant R3.X -> adaptations to the code were done
• transfer code for 2 variables -> creates model script, visualisation script, metadata schema (not fully tested)
• Gropin
  • look into 1 variable bugs:
  • note of gropin 'notused' is made into a variable - FIXED
  • one set of parameters dont need to be turned into dataframe - FIXED
  • general check of visualisation - DONE
  • objective & description fields fill in with info from Matthias - DONE
  • coefficients should be hardcoded, not parameters of fsk! - DONE
  • list of authors incomplete - FIXED
  • Mode_time2multiply: lagTime, logIncrease - DONE
  • Mode_kinetic: logN0, lagTime, logNEnd, simTime - DONE
  • implemented Baranyi model version of gropin - DONE
  • developed knime workflow for creating fskx files from script and md schema
  • wrote editing script for fskx files to fix Creation Date & Reference Bug of FSK Creator
  • uploaded 1-&2-Parameter growth models to server

April

• showcases gropin 24,492,256 done -> concept rewrite for creating full transfer code

30/3

• Lindqvist Telco

• finish Lindqvist model

  29/3

• updated project TODO list

• Taras PhD

• Lindqvist model adaptations and updates

• finishing RRM check

• finishing DRM check

• gropin

  • showcase update
  • transfer annotation and update to do list
  • transfer cleanup
  • update todo

    22/3 - 26/3

• update/orientation of state of things

• created workflow for trasferring gropin model output to javascript plotting node

• output for RRM standalone

• check Petras work for CPM/DRM/RRM updates

• created workflow for efficient testing of CPM/DRM/RRM updates

• testing new FSK build

DONE 2020

11/11/20

• write report for EFSA, as short as possible
• double check reference and authorship for all models
• reply to Panos and thank him and inform about Elternzeit
• reply to Cristina about Possas and include Petra into loop, Elternzeit
• contact Maarten for support, Elternzeit

  10/11

• report for EFSA
  • send to Esther/Matthias -> feedback? ... done
• reply to ahmad ... done
• reply to Panos and thank him and inform about Elternzeit...done
• reply to Cristina about Possas and include Petra into loop, Elternzeit... done
• contact Maarten for support, Elternzeit... done
• share current state of FIID paper with MF
• Lindqvist
  • fix visualisation as newly introduced parameters are not shown
• Gropin
  • annotate code

    09/11

• report for EFSA
  • coordinate with esther...done
  • collaboration with which people?... done
  • table as querformat...done
• Video tutorial support
  • support Petra on video tutorials...done
  • tech sound support for Esther and Petra ... done

    03/11

• Possas
  • bugfixing
• gropin
  • visualisation for 4 or more parameters and therefore finishing growth models
  • webcall with Panos
• support Petra on supply chain models

  02/11

• implement first part of Cristinas Model, further called Possas Model
• tech test of new recording equipment
• gropin
  • visualisation for 4 or more parameters and therefore finishing growth models
  • webcall with Panos

    30/10

• EFSA MRA network meeting

  29/10

• started working on visualisation for 4 or more parameters and therefore finishing growth models
• testing Ahmads new FSK-version
• preparing for EFSA MRA network meeting

  28/10

• gropin
  • showcase models adapted by hand
  • work plan edited with new requests
• supply chain models
  • supported Petra on modelling efforts

    27/10

• gropin
  • showcase models adapted by hand

    26/10

• gropin
  • schedule webcall with Panos
  • rewriting showcase models to run without including a library
  • bugfixing automatic transfer script for all growth models of 3 or less variables
• cristinas model
  • initial discussion on paper and code
• support Petra on supply chain models

  23/10

• gropin
  • bugfixing automatic transfer script for all growth models of 3 or less variables

    22/10

• support Petra on supply chain models
• gropin
  • asking developers about fsk-lab behaviour(Editor deletes simulations that are not default) -> opened a new ticket
  • bugfixing showcase transfer script... DONE
  • questions for devs:
    • MD schema editing data type for parameter DOUBLE or what? ... Answered
    • MD schema date format? ... pending

      21/10

• gropin
  • bugfixing metadata schema editing
  • visualisation for 4 parameters approach, but still buggy

    20/10

• Lindqvist
  • finish adaptions to fskx file
  • write answer
• gropin
  • edit uploaded model, done but due to unexpected fsk-lab behaviour, cannot be edited properly -> ask developers!
  • write questions for gropin creator
  • bugfixing transfer script
• support Petra on supply chain models

  19/10

• support Petra on supply chain models
  • concept plan for hdf5 input of model
  • discussed concept of super/meta nodes network for product conversion
• Lindqvist Model
  • handling comments from lindqvist

    16/10

• gropin
  • finishing showcase model
  • bugfixing transfer script: metadata information ist transferred from gropinDB

    15/10

• gropin
  • creating showcase model with all 3 modes: response surface, time2multiply, kinetic model
  • bugfixing transfer script

    14/10

• gropin
  • bugfixing transfer script (creating 360 annotation schemas, modelscripts and visualisation scripts) for response surface model
  • created knime workflow for taking these 360*3 fsk model components to build fskx models
• support Petra on supply chain models

  13/10

• gropin
  • implementing conditions for growth models with 1,2 variables
  • bugfixing model creation for 3-variable-models
  • found non-functioning models, compiling a list of models not ready for transfer
  • ongoing: checking conditions for models with 4 variables and more... done
  • finished first draft of creating model scripts, visualisation scripts and metadata schema sheets for fskx -> next step: write knime workflow for creator node to create fskx files!

    12/10

• gropin
  • reproducing kinetic models failed
  • model equations are not shown in excel DB, probably hard coded into macros
  • looked into reference given in manual, recreated kinetic model, values are similar but not a match (differences 10-50%)
  • -> need to look into macro coding itself -> problem with BfR's "no macro!" policy -> Lars suggests to use VK-notebook
• supply chain models
  • support for expanding network model

    09/10

• gropin
  • change of plans after first tests with knime -> simpler to create an R script to transfer all gropin models to R
  • successfully extracted data from gropin DB and ordered all models by type:
  • growth models
  • growth / no growth models
  • gamma models with interaction
  • inactivation models
  • lethality models
  • integrated models not included, they seem to need data -> investigation for later
  • findings:
  • every model type needs their own transfer script!
  • within one model type, there several different applications and models in itself
  • starting small: applying transfer script to growth models and concentrating on response surface model application
  • kinetic model application (differential equation) has some parameters that seem to be hardcoded into the macro of gropin, no documentation available, maybe need to investigate in publications
  • finished response surface growth model with 3 parameters
  • problems with
    • 1 parameter -> different plot type
    • 2 parameters -> no subset to choose
    • 4 or more parameters -> subset to choose is more than 1D

08/10

• gropin
  • transfer analysis -> start with 1 group of models and only 1 simple part of the modeling: response surface
  • plan for adopting a 1st part of the models:
  • creating knime workflow for creating R script
    • extract equations and parameters of GRT (growth models) and create response surface models
    • model script contains transformed equation
    • visualisation script contains a switch which response surface should be shown
  • creating knime workflow for editing metadata schema
    • check if all mandatory fields are filled
    • parameters!
    • in a 2nd iteration fill all optional fields with information from database
  • next steps for GRT: kinetic models contain parameters, that are not explained in the user guide, maybe i'll find something in the corresponding papers

07/10

• supply chain models
  • created roadmap for Petra
  • model code review
• Lindqvist model
  • review of comments
• testing recording equipment

06/10

• supply chain models
  • support Petra
• Gropin
  • assessment and transfer analysis

05/10

• Gropin
  • assessment and transfer analysis

02/10

• Taenia model finished, sent Arno related mail

01/10

• Taenia model
  • continuing analyis and transfer assessment
  • code adaptations done --> functioning fskx file
  • annotation

30/09

• supporting Petra with
  • testing new FSK Lab version
  • lindqvist model rewrites
• supporting Lars with Knime server python issues

29/09

• addressing issues of RAKIPInitiative, testing, testing, testing

28/09

• Taenia model
  • code testing and transfer effort assessment
• tubs model
  • code testing and transfer assessment -> no easy way of integrating COMSOL models into fskx, would need to familiarize with COMSOL, then rewrite of models -> estimate 10-20d of work and access to COMSOL license
• jukka models and openBUGS
  • found openBUGS problem, 2 jobs using openBUGS running on the server simultaneously: reproducable problem -> Lars
• bugfixing lindqvist model
  • minor update to match image output with publication format

25/09

• jukka model
  • tests on internal server: fine
  • tests on external server: weird behaviour with openBUGS communication between knime, R and openBUGS - investigating
  • contacted Jukka and told him about models to check out

24/09

• jukka model
  • finished annotating

23/09

• fixing minor mistakes in DRM metadata schema
• jukka model
  • test on server - failed, investigating
• general tests on VRE

22/09

• jukka microbial criterion model
  • annotate

21/09

• nauta
  • 3 example models CPM/DRM combinations
  • annotate
  • test on server

18/09

• jukka microbial criterion model
  • transfer R code to fskx
  • annotate

17/09

• nauta
  • 3 example models CPM/DRM combinations
  • copy&paste R code into fskx
  • test functionality
  • annotate
  • test on server

10/09

• lindqvist Listeria
  • visualisation -> table as in results2.xlsx ... DONE
  • output of xlsx -> remove ... DONE
  • contact lindqvist with Maarten in cc for support ... DONE
  • test on server: ... DONE
  • check which packages are needed after rewrite of code ... DONE
  • tell Lars which packages need to be installed ... DONE

09/09

* rewrite code for fskx transformation ... DONE
  * rewrite m.g.QMRA.R as function inside master.R script
* Define input parameters ... DONE
  * number of runs
  * population group

08/09

• nauta efsa opinion
  • review and test of all CPMs, DRMs and RRM fskx files (execution and annotation) ... in progress
    • adapting DRM module versions with explanation ... in progress
  • test CPMs on server ... DONE
  • implement Maartens comments ... in progress
    • table for quantile results as an optional visualisation ... DONE
    • insert title (CPM/DRM/MS) ... DONE
    • transform probabilities into percentage values with 1 decimal ... DONE
    • random CPM/DRM combination (for every MC iteration, may need to restructure code) ... REJECTED (until further notice)
• lindqvist Listeria
  • analysed R code ... DONE
  • test run ... DONE, success
  • rewrite code for fskx transformation ... in progress
    • rewrite m.g.QMRA.R as function inside master.R script
    • Define input parameters
    • number of runs
    • population group
    • visualisation -> table as in results2.xlsx
    • output of xlsx -> remove
    • test on server:
    • check which packages are needed after rewrite of code
    • tell Lars which packages need to be installed

31/08 - 01/09

• review and test of all CPMs, DRMs and RRM fskx files (execution and annotation) ... in progress
  • CPMs insert author list ... DONE
  • CPMonly versions? remove unnecessary input/output parameters ... REJECTED
  • DRMs insert author list ... DONE
  • adapting DRM module versions with explanation ... in progress
  • RRM insert author list
• test CPMs on server as soon as CPM is ready to be tested ... in progress
  • DRMs can be uploaded and run properly, but encounters problems in FSK-Web, all models run on server
• implement Maartens comments ... in progress
  • table for quantile results as an optional visualisation
    • insert title (CPM/DRM/MS)
    • transform probabilities into percentage values with 1 decimal
  • random CPM/DRM combination (for every MC iteration, may need to restructure code)

28/08

• distribute new visualisation code to all CPMs and DRMs ... in progress
  • finishing CPMs ... DONE
  • finishing DRMs ... in progress
    • creating DRMonly versions ... DONE
    • adapting DRM module versions with explanation: low priority since joiner doesnt work properly
  • finishing RRM ... in progress
    • visualisation errorbars ... DONE
    • random CPM/DRM/MS... DONE
    • test in FSK-Lab ... DONE
• test CPMs on server as soon as CPM is ready to be tested ... pending, waiting for developers
• finalise RRM annotation (need some input see below) ... DONE
• fix a number of known errors in CPM/DRM annotation ... DONE
• review and test of all CPMs, DRMs and RRM fskx files (execution and annotation) ... in progress
  • CPMs ... DONE
  • DRMs ... pending, waiting for Petras review
  • RRM ... DONE
• create a few example CPM/DRM combination as fskx models ... in progress

27/08

• distribute new visualisation code to all CPMs and DRMs ... in progress
  • finishing CPMs ... in progress
  • finishing DRMs ... in progress
    • creating DRMonly versions ... in progress
    • adapting DRM module versions with explanation: low priority since joiner doesnt work properly
  • finishing RRM ... in progress
    • visualisation errorbars
    • random CPM/DRM/MS
    • test in FSK-Lab
• test CPMs on server as soon as CPM is ready to be tested ... pending, waiting for developers
• finalise RRM annotation (need some input see below) ... in progress
• fix a number of known errors in CPM/DRM annotation ... DONE
• review and test of all CPMs, DRMs and RRM fskx files (execution and annotation) ... in progress
  • CPMs ... in progress
  • DRMs ... pending, waiting for Petras review
  • RRM ... in progress
• create a few example CPM/DRM combination as fskx models

26/08

• distribute new visualisation code to all CPMs and DRMs ... in progress
  • finishing CPMs ... DONE
  • finishing DRMs ... pending, waiting for Maartens comments
  • finishing RRM ... pending, waiting for Maartens comments
  • test CPMs on server as soon as CPM is ready to be tested ... pending, waiting for developres
• finalise RRM annotation (need some input see below) ... in progress
• fix a number of known errors in CPM/DRM annotation ... DONE
• review and test of all CPMs, DRMs and RRM fskx files (execution and annotation) ... in progress
• create a few example CPM/DRM combination as fskx models

25/08

• clean up R code ... in progress
  • CPMs ... DONE
  • DRMs ... DONE
  • RRM ... DONE
• distribute new visualisation code to all CPMs and DRMs ... in progress
  • finishing CPMs ... DONE
  • finishing DRMs ... pending, waiting for Maartens comments
  • finishing RRM ... pending, waiting for Maartens comments
• cleaning up obsolete code snippets and output parameters to increase transparency ... DONE
• test CPMs on server as soon as CPM is ready to be tested ... not possible at the moment, Thomas is investigating
• finalise RRM annotation (need some input see below) ... in progress
• fix a number of known errors in CPM/DRM annotation

24/08

• clean up R code ... in progress
  • CPMs ... DONE
• distribute new visualisation code to all CPMs and DRMs ... in progress
  • finishing CPMs ... mostly done, mylius missing
• cleaning up obsolete code snippets and output parameters to increase transparency
• test CPMs on server as soon as CPM is ready to be tested ... in progress, stuck at uploading -> asking Lars/Thomas for help

21/08

• implement visualisation fixes for CPMs
  • find out about reason for steplike function at low doses ... DONE
  • fix cdf should increase to 1 ... DONE
• implement visualisation fixes for DRMs ... DONE
• implement visualisation with all reduction (or "intervention") scenarios ... DONE
• implement RRM standalone fskx model with all 24 CPM/DRM-combinations as a csv table input and have the user choose CPM and DRM as input parameter ... DONE

Lindqvist EFSA opinion

Reference

doi: 10.2903/j.efsa.2020.6092

Questions for Maarten - call 2

• visualisation of CPM ... fine, according to Maarten, confirmation?
  • Maarten: d
• visualisation of DRM -> changes, more infos?
  • Maarten:
• visualisation of RRM
  • may show all intervention scenarios at the same time (load from table) = same vis as in @Risk -> better?
  • IF ONLY one intervention scenario: should the name (FA1/2/3...) be put into visualisation?
  • which text infos in visualisation?
  • Maarten: all intervention plus one from user
• annotation of RRM:
  • Inputs from MS skin results
    • definition?
    • how to annotate?
    • EU values are weighted average of all MS, here is just taken as input value, not calculated -> problem?
  • meat conc distributions Input scale (CFU skin) as parameter or leave it as is? how to annotate?
  • slope - definition? or where to find
  • tau - definition? or where to find
  • delta fex - definition? or where to find

Questions for Maarten - call 1

general questions model related

• Cret in R code is randomly chosen (normal distribution), in @Risk as abscissa with certain range for Pill
  • -> new model?s
  • Maarten: OK
• new visualisation dose vs cret --> possible to simulate the other plot, sampling from Cret-axis and get dose distribution
  • possible to recreate old visualization -> would that be better?
  • Maarten: YES
  • (current visualization RAKIP models: what is ordinate? what is info i get out of it?) Distribution given Cret
• Maxportion in RAKIP = 1000, but in @Risk no upper bound is given... what to do?
  • Maarten: take Maxportion = 1000
• in R models is Prev (probability of prevalence) used --> also for efsa opinion relevant?
  • Maarten: use it in code as is
• dose in RAKIP models only binomial distribution, dose in @Risk model is normal -> which one ?
  • Maarten: BOTH (for either case)

specific CPM related questions

• Lindqvist: Ncarcass vs Nportion?
  • Maarten: summ Nportion to Ncarcass (1097 average carcass size)
• Van Asselt: washed board prob =0, but unwashed board depends on hands => shouldnt washed board with unwashed hands do something?
  • Maarten: its ok
• calistri model: empiric transfer probability hands to meat 2x 0.6? is that correct? its a cdf, isnt it?
  • Maarten: Mistake;
  • Marcel checked in paper: 0.8 is correct, corrected in R code, but differences are minor <1%

regression model questions

• in Campy regression @Risk code in tab "meat conc distributions", column log dose interval : first and last values weird source cell in excel -> better?
  • Maarten: reason is matching pattern in delta -> Marcel simulates this in R code (even if changes would be minor)

general R problems

• R 3.X: some functions(rbinom, gamma, etc.) have only a 32bit range of numbers --> some models use numbers outside of 32bit range - solved in R4 but FSK only supports R3.X

Christensen et al.: Risk Assessment of Campylobacter jejuni in Chicken Products

status

• found fskx file in RAKIP model repository, not fully annotated, discrepancies between code and reference
• refers to DOI: 10.1111/j.1539-6924.2010.01481.x --> checking that reference led to similar publication

  todo

• need to check if model code and code in reference are identical
• annotation completing

  Reference

  Christensen B, Sommer H, Nielsen N and Rosenquist H, 2001. Risk Assessment of Campylobacter jejuni in Chicken Products. Danish Veterinary and Food Administration, Copenhagen, Denmark.

Brynestad et al.: Quantitative microbiological risk assessment of campylobacteriosis cases in the German population due to consumption of chicken prepared in homes

status

• found fskx file in RAKIP model repository, not fully annotated, discrepancies between code and reference
• refers to DOI: 10.1111/j.1539-6924.2010.01481.x --> checking that reference led to similar publication

  todo

• need to check if model code and code in reference are identical
• annotation completing

  Reference

  Brynestad S, Braute L, Luber P and Bartelt E, 2008. Quantitative microbiological risk assessment of campylobacteriosis cases in the German population due to consumption of chicken prepared in homes. International Journal of Risk Assessment and Management, 8, 194–213

Nauta et al.: Food safety in the domestic environment: the effect of consumer risk information on human disease risks

status

• found fskx file in RAKIP model repository, not fully annotated, discrepancies between code and reference
• refers to DOI: 10.1111/j.1539-6924.2010.01481.x --> checking that reference led to DIFFERENT publication

  todo

• need to check if model code and code in reference are identical
• annotation completing

  Reference

  Nauta MJ, Fischer AR, van Asselt ED, de Jong AE, Frewer LJ and de Jonge R, 2008. Food safety in the domestic environment: the effect of consumer risk information on human disease risks. Risk Analysis, 28, 179–192.

WHO: Risk assessment of Campylobacter spp. in broiler chickens: technical report

status

• found fskx file in RAKIP model repository, not fully annotated, discrepancies between code and reference
• refers to DOI: 10.1111/j.1539-6924.2010.01481.x --> checking that reference led to DIFFERENT publication

  todo

• need to check if model code and code in reference are identical
• annotation completing

  Reference

  WHO, 2009. Risk assessment of Campylobacter spp. in broiler chickens: technical report. Microbiological Risk Assessment Series, 12

CPMs currently in process of transfer to fskx

Mylius et al.: Cross-contamination during food preparation: a mechanistic modelapplied to chicken-borne Campylobacter

Reference

Mylius SD, Nauta MJ and Havelaar AH, 2007. Cross-contamination during food preparation: a mechanistic model applied to chicken-borne Campylobacter. Risk Analysis, 27, 803–813.

van Asselt et al.: Cross-contamination in the kitchen: estimation of transfer rates for cutting boards, hands and knives

Reference

van Asselt ED, de Jong AE, de Jonge R and Nauta MJ, 2008. Cross-contamination in the kitchen: estimation of transfer rates for cutting boards, hands and knives. Journal of Applied Microbiology, 105, 1392–1401.

Calistri and Giovannini: Quantitative risk assessment of human campylobacteriosis related to the consumption of chicken meat in two Italian regions

Reference

Calistri P and Giovannini A, 2008. Quantitative risk assessment of human campylobacteriosis related to the consumption of chicken meat in two Italian regions. International Journal of Food Microbiology, 128, 274–287.

Lindqvist and Lindblad: Quantitative risk assessment of thermophilic Campylobacter spp. and crosscontamination during handling of raw broiler chickens evaluating strategies at the producer level to reduce human campylobacteriosis in Sweden

Reference

Lindqvist R and Lindblad M, 2008. Quantitative risk assessment of thermophilic Campylobacter spp. and crosscontamination during handling of raw broiler chickens evaluating strategies at the producer level to reduce human campylobacteriosis in Sweden. International Journal of Food Microbiology, 121, 41–52.

DRM

3 dose response models are compared to each other

Jukka Ranta models

Jukka Ranta et al.: unnamed Dose response model

Jukka Ranta et al.: A Bayesian approach to the evaluation of risk-based microbiological criteria for Campylobacter in broiler meat

https://projecteuclid.org/download/pdfview_1/euclid.aoas/1446488745

outlook

EFSA / Lindqvist et al.: mgQMRA Listeria monocytogenes in frozen fruit, vegetables and herbs

Reference

https://www.efsa.europa.eu/en/efsajournal/pub/6092

Meester / Swart et al.: A quantitative risk assessment for human Taenia solium exposure from home slaughtered pigs in European countries

Deng / Swart et al: Quantitative risk assessment of meat-borne Toxoplasma gondii infection in the mainland of China

Swart et al: A quantitative risk model for Trichinella spp. in pork and wild boar meat

Teunis et al: Acute illness from Campylobacter jejuni may require high doses while infection occurs at low doses

Source attribution models (microbial subtyping for salmonellosis) – work funded by MATRIX project

Reference

https://onlinelibrary.wiley.com/doi/full/10.1111/zph.12645

EFSA: tubs model → certain elements might be possible (growth model) KJ

Other ideas

• https://www.sciencedirect.com/science/article/abs/pii/S0360132320303292
• Campylobacter DRM: https://onlinelibrary.wiley.com/doi/full/10.1111/risa.12572