Integrating Crop Growth Model and Constraint-Based - PowerPoint Presentation

Slide1

Integrating Crop Growth Model and Constraint-BasedMetabolic Model

Maurice Cheung

9 June 2021

Slide2

AcknowledgementsOrganisers of Crops

in silico

Rahul ShawMy former postdocYale-NUS Collegea small liberal arts college in Singaporecollaboration between Yale and NUS (National University of Singapore)

primarily undergraduate institution (PUI)

Slide3

Crop growth model x Constraint-based metabolic model

Allow the predictions of metabolic flux patterns under different environmental conditions

Shaw and Cheung, unpublished

Slide4

Construction of genome-scale metabolic modelsDNA sequences of an organism including all of its genesFunctional annotations of genes in the organismExtract metabolic enzymes from annotated genomes

Metabolic reactions catalysed by the enzymesMathematic model containing 100s – 1000s reactions and metabolites

Slide5

Genome-scale metabolic models

Roche Applied Science "Biochemical Pathways" wall chart; http://web.expasy.org/pathways/

Slide6

Gene-Protein-Reaction associations

Genome-scale models contains information on gene-protein-reaction associations

allows integration of

transcriptomic

and

proteomic

data

Proteome

Transcriptome

Metabolic network

Slide7
Genome-scale models of crop plants

Maize (2010; 2011)Rice (2014; 2015)Tomato (2016)

Soybean (2019)

Poolman

, M. G.,

Kundu

, S., Shaw, R., & Fell, D. A. (2014). Frontiers in Plant Science, 5, 1–7.

Lakshmanan

, M., Lim, S.-H.,

Mohanty

, B., Kim, J. K., Ha, S.-H., & Lee, D.-L. (2015). Plant Physiology, 169, pp.15.01379

Dal’Molin

, C. G.,

Quek

, L.-E. L.-E.,

Palfreyman

, R. W.,

Brumbley

, S. M., Nielsen, L. K., (2010). Plant Physiology, 154(4), 1871–85.

Saha

, R.,

Suthers

, P. F., &

Maranas

, C. D. (2011).

PLoS ONE, 6(7), e21784.

Yuan, H., Cheung, C. Y. M., Poolman, M. G., Hilbers, P. A. J., & van Riel, N. A. W. (2016). The Plant Journal, 85, 289–304. Moreira, T.B., Shaw, R., Luo, X., Ganguly, O., Kim, H.-S., Coelho, L.G.F., Cheung, C.Y.M., Williams, T.C.R. (2019). Plant Physiology, 180, 1912-29.

Slide8

Curation of rice genome-scale metabolic model:senescence and vitamin synthesis

To model resource remobilisationDegradation of biomassNeed to curate degradation pathwaysEspecially chlorophyll degradationAlso amino-acids etc.

Vitamin

in seed biomass

B2, B3, B5, B6 and folate

Curate for vitamin synthesis pathways

Mass- and charge-balancing

Energy conservation

Gene-protein-reaction associations

Connect with transcriptomes and proteomes

Wang et al. (2020) Nature Communications, 11, 1254

Slide9

FBA is the most commonly used modelling technique for analysing genome-scale metabolic models

Constraint-based modelling:

Flux balance analysis (FBA)

Carbohydrate

Protein

Lipid

Nucleic acids

Pi

SO

4

2-

NO

3

-

CO

2

Slide10

Limitations of constraint-based metabolic modellingNot easy to incorporate

environmental factorsWater availabilityTemperatureThese factors are affected by climate change

Töpfer et al. (2020) The Plant Cell, 32(12), 3689-3705

Gas exchange model x constraint-based model

IPCC AR5 (2014) Figure SPM.7

Slide11

Crop growth model – WOFOSTMechanistic crop growth model

Relate environmental factors to crop growthWOrld FOod STudies (WOFOST) model

Incorporate environmental factorslight, temperature, water and macro-nutrients (N, P, K)Predict crop growthSeed, leaf, stem

de Wit et al. (2019) Agricultural Systems, 168, 154-167

Slide12

Rice organ growth predicted by WOFOSTWOFOST model outputs:

biomass of leaf, stem, seed

vegetative, reproductive, grain filling stagesnormal and water-limited (stress) conditions

asdsad

Shaw and Cheung, unpublished

Slide13

Using organ growth / degradation rates as constraints for flux balance model

Calculate

change in biomass based on WOFOST outputUse change in biomass as constraints to FBA modelleaf, stem, seedper dayDifferent constraints for normal and water-limited conditions

Shaw and Cheung, unpublished

Slide14

Multi-organ rice genome-scale metabolic model3 organs –

leaf, stem, seedConnected by common pool (CP)phloem and xylemEach organ has its own biomass equations

Can access N, S, P from CP Effectively from rootOnly leaf received photonsDiel modelling framework

Shaw and Cheung, unpublished

Slide15

Set up FBA model for normal & water-limited stress

Biomass constraints as the only difference between normal and water-limited conditions

Run simulations day-wise for 100 daysVegetative, reproductive, grain filling stages

Shaw and Cheung, unpublished

Slide16

Predictions from FBA modelCompare between normal and water-limited conditions

Photosynthesissimilar at vegetative stagelevel off earlier for water-limitedN assimilation

low at the end of grain filling under normal conditionincrease at the end of grain filling in water-limited condition

Shaw and Cheung, unpublished

Slide17

Transport of amino-acids for grain fillingGly

, Arg, Ser are the most efficient amino-acids as N source to the seed under both conditionsPredicted from FBA model

Patterns of amino-acids transport to the seed are very different between two conditionslargely follow the pattern of seed development

Shaw and Cheung, unpublished

Slide18

Metabolic changes across developmental stages and environmental conditions

Number of unique reactions in each “stage-condition”Many unique reactions in grain-filling stage

under water-limited stressVery little difference between 2 conditions during vegetative stage

Shaw and Cheung, unpublished

Slide19

Metabolic interactions between organs: Resource (re-)allocation

Flux solutions from FBA allow us to study metabolic flux distributionsPattern of resource allocation is similar between normal and water-limited conditions

Leaf is the main source organ supplying sucrose and amino-acids to the other organs

Shaw and Cheung, unpublished

Slide20

Future of crop plant modelling: Multi-scale modelsMany

models at different scales developed and being perfected over decadesThere are limitations /

assumptions made for each modelling frameworkIntegrate models of different scales to overcome the limitations

For new (and experienced) modellers:

Have expertise in at least one type of modellingUnderstand its assumptions and limitationsKnow

how to connect to other types of modelsDon’t be afraid to learn new types of modelsNeed to know enough about other types of model

Figure out the

connection points

between models

Remember to

enjoy the process of modelling

!

good luck; have fun :)

Shaw and Cheung, unpublished