Legacy Bug ID: 1655
Author: Scott
Date: 2013-02-27 12:21:54 +0000 UTC
STEP 1: PROPOSAL
Short description of the concept or proposed improvements (describe the problem):
APSIM wheat flowering time is parameterised by VRN, PPD and thermal time coefficients and there are multiple solutions for any given genotype. we address 2 issues (1) the VRN and PPD effects are considered as Maximum effects, which doesnt properly deal with interactions and (2) plant breeders are interested in flowering time as controlled by known genes (gene-based model)
Rationale and Justification (why is this problem important):
We want to try using the model in this way to make it relevant for prediction of 'virtual genotypes' and applications in plant breeding.
The gene-based part of the implementation does not have to be in any release, but we regard this as a useful development version to work with plant breeders.
Details of the proposed improvements (how you want to fix it):
(1) implementing a multiplicative function for the Fv and Fp parameters, i.e. TT = Sigma(deltaTT x Fv x Fp), and extend the photoperiod sensitivity period from floral initiation to flowering time. (2) implement a gene-based parameterisation of APSIM-Wheat whereby flowering time is controlled by the number of VRN alleles and PPD alleles and a residual thermal time (earliness per se). All is detailed in manuscript, and has been implemented and tested using the APSIM routines, extracted into the R language
Implications for users (how will this particular fix help):
We validated the predictions on a large dataset (>5000 observations), and we also 'refitted' all of the genotypes that are in the APSIM release, using the original APSIM algorithms. We found that some of those genotypes in the release are poorly parameterised and should be updated. The implementation of gene-based parameters allows the evaluation of 'novel' wheat genotypes based on their allele composition for major VRN and PPD genes. This is known for most commercial genotypes. The gene-based parameterisation facilitates a much simpler validation process - we showed that we only need a single sowing with +/- vern and +/- photoperiod treatments in order to calibrate flowering time parameters.
Legacy Bug ID: 1655 Author: Scott Date: 2013-02-27 12:21:54 +0000 UTC
STEP 1: PROPOSAL
Short description of the concept or proposed improvements (describe the problem):
APSIM wheat flowering time is parameterised by VRN, PPD and thermal time coefficients and there are multiple solutions for any given genotype. we address 2 issues (1) the VRN and PPD effects are considered as Maximum effects, which doesnt properly deal with interactions and (2) plant breeders are interested in flowering time as controlled by known genes (gene-based model)
Rationale and Justification (why is this problem important): We want to try using the model in this way to make it relevant for prediction of 'virtual genotypes' and applications in plant breeding. The gene-based part of the implementation does not have to be in any release, but we regard this as a useful development version to work with plant breeders.
Details of the proposed improvements (how you want to fix it):
(1) implementing a multiplicative function for the Fv and Fp parameters, i.e. TT = Sigma(deltaTT x Fv x Fp), and extend the photoperiod sensitivity period from floral initiation to flowering time. (2) implement a gene-based parameterisation of APSIM-Wheat whereby flowering time is controlled by the number of VRN alleles and PPD alleles and a residual thermal time (earliness per se). All is detailed in manuscript, and has been implemented and tested using the APSIM routines, extracted into the R language
Implications for users (how will this particular fix help):
We validated the predictions on a large dataset (>5000 observations), and we also 'refitted' all of the genotypes that are in the APSIM release, using the original APSIM algorithms. We found that some of those genotypes in the release are poorly parameterised and should be updated. The implementation of gene-based parameters allows the evaluation of 'novel' wheat genotypes based on their allele composition for major VRN and PPD genes. This is known for most commercial genotypes. The gene-based parameterisation facilitates a much simpler validation process - we showed that we only need a single sowing with +/- vern and +/- photoperiod treatments in order to calibrate flowering time parameters.
Reviewer:?
Review Panel Contact: Karine Chenu Supporting documentation (optional): manuscript attached.