Limits...
Inferring the gene network underlying the branching of tomato inflorescence.

Astola L, Stigter H, van Dijk AD, van Daelen R, Molenaar J - PLoS ONE (2014)

Bottom Line: With the network inferred by our algorithm, we can explain the dynamics corresponding to all three genotypes simultaneously, despite their apparent dissimilarities.We also correctly predict the chronological order of expression peaks for the main hubs in the network.Based on the inferred network, using optimal experimental design criteria, we are able to suggest an informative set of experiments for further investigation of the mechanisms underlying branching behavior.

View Article: PubMed Central - PubMed

Affiliation: Biometris, Wageningen University and Research Centre, Wageningen, The Netherlands; Netherlands Consortium for Systems Biology, Amsterdam, The Netherlands.

ABSTRACT
The architecture of tomato inflorescence strongly affects flower production and subsequent crop yield. To understand the genetic activities involved, insight into the underlying network of genes that initiate and control the sympodial growth in the tomato is essential. In this paper, we show how the structure of this network can be derived from available data of the expressions of the involved genes. Our approach starts from employing biological expert knowledge to select the most probable gene candidates behind branching behavior. To find how these genes interact, we develop a stepwise procedure for computational inference of the network structure. Our data consists of expression levels from primary shoot meristems, measured at different developmental stages on three different genotypes of tomato. With the network inferred by our algorithm, we can explain the dynamics corresponding to all three genotypes simultaneously, despite their apparent dissimilarities. We also correctly predict the chronological order of expression peaks for the main hubs in the network. Based on the inferred network, using optimal experimental design criteria, we are able to suggest an informative set of experiments for further investigation of the mechanisms underlying branching behavior.

Show MeSH

Related in: MedlinePlus

Box-plots of estimated parameters.In panel A is a box-plot of the distribution of the successive optimal parameters during the thinning procedure. The values have consistent signs and narrow range. In panel B are the distributions of successive optimizations using the inferred network structure, starting from different random initial guesses with the Matlab routine lsqnonlin, showing again a narrow range of deviation. In panel C we see the evolution of the AFPE-value during the thickening-thinning procedure. We stopped before the last step, where AFPE increases slightly.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3974656&req=5

pone-0089689-g004: Box-plots of estimated parameters.In panel A is a box-plot of the distribution of the successive optimal parameters during the thinning procedure. The values have consistent signs and narrow range. In panel B are the distributions of successive optimizations using the inferred network structure, starting from different random initial guesses with the Matlab routine lsqnonlin, showing again a narrow range of deviation. In panel C we see the evolution of the AFPE-value during the thickening-thinning procedure. We stopped before the last step, where AFPE increases slightly.

Mentions: The final fit has thus 2 additional edges and 5 removed edges compared to the original configuration. For the evolution of the AFPE, throughout the thickening and thinning steps, see Figure 4, panel C. Note that throughout the thickening-thinning procedure, we have simultaneously fitted both data, wild type and mutant, which have rather different dynamics, using the same set of parameters with only the special parameter accounting for the differences. The value of steadily converged to around 0.5, indicating that the influence of the S-gene is 50% weaker in the mutant compared to the wild type. Note that we use global non-constrained optimization without any fixed initial points. Nevertheless the signs of the parameters remained consistent throughout the iteration. For a box-plot of the remaininig optimal parameters during the thinning phase see Figure 4, panel A. As a result, we obtained the minimal network in panel B of Figure 1 that is able to describe the data well. This network contains as many edges as is needed to fit the data, but removing any of them will result in a very poor fit. The algorithm not only unravels which interactions are necessary, but also whether it is a promoting or inhibiting one.


Inferring the gene network underlying the branching of tomato inflorescence.

Astola L, Stigter H, van Dijk AD, van Daelen R, Molenaar J - PLoS ONE (2014)

Box-plots of estimated parameters.In panel A is a box-plot of the distribution of the successive optimal parameters during the thinning procedure. The values have consistent signs and narrow range. In panel B are the distributions of successive optimizations using the inferred network structure, starting from different random initial guesses with the Matlab routine lsqnonlin, showing again a narrow range of deviation. In panel C we see the evolution of the AFPE-value during the thickening-thinning procedure. We stopped before the last step, where AFPE increases slightly.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3974656&req=5

pone-0089689-g004: Box-plots of estimated parameters.In panel A is a box-plot of the distribution of the successive optimal parameters during the thinning procedure. The values have consistent signs and narrow range. In panel B are the distributions of successive optimizations using the inferred network structure, starting from different random initial guesses with the Matlab routine lsqnonlin, showing again a narrow range of deviation. In panel C we see the evolution of the AFPE-value during the thickening-thinning procedure. We stopped before the last step, where AFPE increases slightly.
Mentions: The final fit has thus 2 additional edges and 5 removed edges compared to the original configuration. For the evolution of the AFPE, throughout the thickening and thinning steps, see Figure 4, panel C. Note that throughout the thickening-thinning procedure, we have simultaneously fitted both data, wild type and mutant, which have rather different dynamics, using the same set of parameters with only the special parameter accounting for the differences. The value of steadily converged to around 0.5, indicating that the influence of the S-gene is 50% weaker in the mutant compared to the wild type. Note that we use global non-constrained optimization without any fixed initial points. Nevertheless the signs of the parameters remained consistent throughout the iteration. For a box-plot of the remaininig optimal parameters during the thinning phase see Figure 4, panel A. As a result, we obtained the minimal network in panel B of Figure 1 that is able to describe the data well. This network contains as many edges as is needed to fit the data, but removing any of them will result in a very poor fit. The algorithm not only unravels which interactions are necessary, but also whether it is a promoting or inhibiting one.

Bottom Line: With the network inferred by our algorithm, we can explain the dynamics corresponding to all three genotypes simultaneously, despite their apparent dissimilarities.We also correctly predict the chronological order of expression peaks for the main hubs in the network.Based on the inferred network, using optimal experimental design criteria, we are able to suggest an informative set of experiments for further investigation of the mechanisms underlying branching behavior.

View Article: PubMed Central - PubMed

Affiliation: Biometris, Wageningen University and Research Centre, Wageningen, The Netherlands; Netherlands Consortium for Systems Biology, Amsterdam, The Netherlands.

ABSTRACT
The architecture of tomato inflorescence strongly affects flower production and subsequent crop yield. To understand the genetic activities involved, insight into the underlying network of genes that initiate and control the sympodial growth in the tomato is essential. In this paper, we show how the structure of this network can be derived from available data of the expressions of the involved genes. Our approach starts from employing biological expert knowledge to select the most probable gene candidates behind branching behavior. To find how these genes interact, we develop a stepwise procedure for computational inference of the network structure. Our data consists of expression levels from primary shoot meristems, measured at different developmental stages on three different genotypes of tomato. With the network inferred by our algorithm, we can explain the dynamics corresponding to all three genotypes simultaneously, despite their apparent dissimilarities. We also correctly predict the chronological order of expression peaks for the main hubs in the network. Based on the inferred network, using optimal experimental design criteria, we are able to suggest an informative set of experiments for further investigation of the mechanisms underlying branching behavior.

Show MeSH
Related in: MedlinePlus