A novel batch-effect correction method for scRNA-seq data based on Adversarial Information Factorization

Single-cell RNA sequencing (scRNA-seq) technology produces an unprecedented resolution at the level of a unique cell, raising great hopes in medicine. Nevertheless, scRNA-seq data suffer from high variations due to the experimental conditions, called batch effects, preventing any aggregated downstream analysis. Adversarial Information Factorization provides a robust batch-effect correction method that does not rely on prior knowledge of the cell types nor a specific normalization strategy while being adapted to any downstream analysis task. It compares to and even outperforms state-of-the-art methods in several scenarios: low signal-to-noise ratio, batch-specific cell types with few cells, and a multi-batches dataset with imbalanced batches and batch-specific cell types. Moreover, it best preserves the relative gene expression between cell types, yielding superior differential expression analysis results. Finally, in a more complex setting of a Leukemia cohort, our method preserved most of the underlying biological information for each patient while aligning the batches, improving the clustering metrics in the aggregated dataset.

BR55 Ultrasound Molecular Imaging of Clear Cell Renal Cell Carcinoma Reflects Tumor Vascular Expression of VEGFR-2 in a Patient-Derived Xenograft Model

Standard imaging cannot reliably predict the nature of renal tumors. Among malignant renal tumors, clear cell renal cell carcinoma (ccRCC) is the most common histological subtype, in which the vascular endothelial growth factor 2 (VEGFR-2) is highly expressed in the vascular endothelium. BR55, a contrast agent for ultrasound imaging, consists of gas-core lipid microbubbles that specifically target and bind to the extracellular portion of the VEGFR-2. The specific information provided by ultrasound molecular imaging (USMI) using BR55 was compared with the vascular tumor expression of the VEGFR-2 by immunohistochemical (IHC) staining in a preclinical model of ccRCC. Patients' ccRCCs were orthotopically grafted onto Nod-Scid-Gamma (NSG) mice to generate patient-derived xenografts (PdX). Mice were divided into four groups to receive either vehicle or axitinib an amount of 2, 7.5 or 15 mg/kg twice daily. Perfusion parameters and the BR55 ultrasound contrast signal on PdX renal tumors were analyzed at D0, D1, D3, D7 and D11, and compared with IHC staining for the VEGFR-2 and CD34. Significant Pearson correlation coefficients were observed between the area under the curve (AUC) and the CD34 (0.84, p < 10-4), and between the VEGFR-2-specific signal obtained by USMI and IHC (0.72, p < 10-4). USMI with BR55 could provide instant, quantitative information on tumor VEGFR-2 expression to characterize renal masses non-invasively.

Optimizing IFN Alpha Therapy against Myeloproliferative Neoplasms

Myeloproliferative neoplasms (MPNs) are hematologic malignancies that result from acquired driver mutations in hematopoietic stem cells (HSCs), causing overproduction of blood cells and an increased risk of thrombohemorrhagic events. The most common MPN driver mutation affects the JAK2 gene (JAK2V617F ). Interferon alpha (IFNα) is a promising treatment against MPNs by inducing a hematologic response and molecular remission for some patients. Mathematical models have been proposed to describe how IFNα targets mutated HSCs, indicating that a minimal dose is necessary for long-term remission. This study aims to determine a personalized treatment strategy. First, we show the capacity of an existing model to predict cell dynamics for new patients from data that can be easily obtained in clinic. Then, we study different treatment scenarios in silico for three patients, considering potential IFNα dose-toxicity relations. We assess when the treatment should be interrupted depending on the response, the patient's age, and the inferred development of the malignant clone without IFNα We find that an optimal strategy would be to treat patients with a constant dose so that treatment could be interrupted as quickly as possible. Higher doses result in earlier discontinuation but also higher toxicity. Without knowledge of the dose-toxicity relationship, trade-off strategies can be found for each patient. A compromise strategy is to treat patients with medium doses (60-120 μg/week) for 10-15 years. Altogether, this work demonstrates how a mathematical model calibrated from real data can help build a clinical decision-support tool to optimize long-term IFNα therapy for MPN patients. SIGNIFICANCE STATEMENT: Myeloproliferative neoplasms (MPNs) are chronic blood cancers. Interferon alpha (IFNα) is a promising treatment with the potential to induce a molecular response by targeting mutated hematopoietic stem cells. MPN patients are treated over several years, and there is a lack of knowledge concerning the posology strategy and the best timing for interrupting therapy. The study opens avenues for rationalizing how to treat MPN patients with IFNα over several years, promoting a more personalized approach to treatment.

A biology-driven deep generative model for cell-type annotation in cytometry

Cytometry enables precise single-cell phenotyping within heterogeneous populations. These cell types are traditionally annotated via manual gating, but this method lacks reproducibility and sensitivity to batch effect. Also, the most recent cytometers-spectral flow or mass cytometers-create rich and high-dimensional data whose analysis via manual gating becomes challenging and time-consuming. To tackle these limitations, we introduce Scyan https://github.com/MICS-Lab/scyan, a Single-cell Cytometry Annotation Network that automatically annotates cell types using only prior expert knowledge about the cytometry panel. For this, it uses a normalizing flow-a type of deep generative model-that maps protein expressions into a biologically relevant latent space. We demonstrate that Scyan significantly outperforms the related state-of-the-art models on multiple public datasets while being faster and interpretable. In addition, Scyan overcomes several complementary tasks, such as batch-effect correction, debarcoding and population discovery. Overall, this model accelerates and eases cell population characterization, quantification and discovery in cytometry.

Synergic prognostic value of 3D CT scan subcutaneous fat and muscle masses for immunotherapy-treated cancer

BACKGROUND

Our aim was to explore the prognostic value of anthropometric parameters in a large population of patients treated with immunotherapy.

METHODS

We retrospectively included 623 patients with advanced non-small cell lung cancer (NSCLC) (n=318) or melanoma (n=305) treated by an immune-checkpoint-inhibitor having a pretreatment (thorax-)abdomen-pelvis CT scan. An external validation cohort of 55 patients with NSCLC was used. Anthropometric parameters were measured three-dimensionally (3D) by a deep learning software (Anthropometer3DNet) allowing an automatic multislice measurement of lean body mass, fat body mass (FBM), muscle body mass (MBM), visceral fat mass (VFM) and sub-cutaneous fat mass (SFM). Body mass index (BMI) and weight loss (WL) were also retrieved. Receiver operator characteristic (ROC) curve analysis was performed and overall survival was calculated using Kaplan-Meier (KM) curve and Cox regression analysis.

RESULTS

In the overall cohort, 1-year mortality rate was 0.496 (95% CI: 0.457 to 0.537) for 309 events and 5-year mortality rate was 0.196 (95% CI: 0.165 to 0.233) for 477 events. In the univariate Kaplan-Meier analysis, prognosis was worse (p<0.001) for patients with low SFM (<3.95 kg/m2), low FBM (<3.26 kg/m2), low VFM (<0.91 kg/m2), low MBM (<5.85 kg/m2) and low BMI (<24.97 kg/m2). The same parameters were significant in the Cox univariate analysis (p<0.001) and, in the multivariate stepwise Cox analysis, the significant parameters were MBM (p<0.0001), SFM (0.013) and WL (0.0003). In subanalyses according to the type of cancer, all body composition parameters were statistically significant for NSCLC in ROC, KM and Cox univariate analysis while, for melanoma, none of them, except MBM, was statistically significant. In multivariate Cox analysis, the significant parameters for NSCLC were MBM (HR=0.81, p=0.0002), SFM (HR=0.94, p=0.02) and WL (HR=1.06, p=0.004). For NSCLC, a KM analysis combining SFM and MBM was able to separate the population in three categories with the worse prognostic for the patients with both low SFM (<5.22 kg/m2) and MBM (<6.86 kg/m2) (p<0001). On the external validation cohort, combination of low SFM and low MBM was pejorative with 63% of mortality at 1 year versus 25% (p=0.0029).

CONCLUSIONS

3D measured low SFM and MBM are significant prognosis factors of NSCLC treated by immune checkpoint inhibitors and can be combined to improve the prognostic value.

CustOmics: A versatile deep-learning based strategy for multi-omics integration

The availability of patient cohorts with several types of omics data opens new perspectives for exploring the disease's underlying biological processes and developing predictive models. It also comes with new challenges in computational biology in terms of integrating high-dimensional and heterogeneous data in a fashion that captures the interrelationships between multiple genes and their functions. Deep learning methods offer promising perspectives for integrating multi-omics data. In this paper, we review the existing integration strategies based on autoencoders and propose a new customizable one whose principle relies on a two-phase approach. In the first phase, we adapt the training to each data source independently before learning cross-modality interactions in the second phase. By taking into account each source's singularity, we show that this approach succeeds at taking advantage of all the sources more efficiently than other strategies. Moreover, by adapting our architecture to the computation of Shapley additive explanations, our model can provide interpretable results in a multi-source setting. Using multiple omics sources from different TCGA cohorts, we demonstrate the performance of the proposed method for cancer on test cases for several tasks, such as the classification of tumor types and breast cancer subtypes, as well as survival outcome prediction. We show through our experiments the great performances of our architecture on seven different datasets with various sizes and provide some interpretations of the results obtained. Our code is available on (https://github.com/HakimBenkirane/CustOmics).

Better than RECIST and faster than iRECIST: defining the Immunotherapy Progression Decision score to better manage progressive tumors on immunotherapy

PURPOSE

The objective of the study is to propose the immunotherapy progression decision (iPD) score, a practical tool based on patient features that are available at the first evaluation of immunotherapy treatment, to help oncologists decide whether to continue the treatment or switch rapidly to another therapeutic line when facing a progressive disease patient at the 1st evaluation.

METHODS

This retrospective study included 107 patients with progressive disease at first evaluation according to RECIST 1.1. Clinical, radiological, and biological data at baseline and 1st evaluation were analyzed. An external validation set consisting of 31 patients with similar baseline characteristics was used for the validation of the score.

RESULTS

Variables were analyzed in a univariate study. The iPD score was constructed using only independent variables, each considered as a worsening factor for the survival of patients. The patients were stratified in three groups: Good Prognosis (GP), Poor Prognosis (PP) and Critical Prognosis (CP). Each group showed significantly different survivals (GP: 11.4, PP: 4.4, CP: 2.3 months median OS, p<0.001 log-rank-test). Moreover, the iPD score was able to detect the pseudo-progressors better than other scores. On the validation set, CP patients had significantly worse survival than PP and GP patients (p<0.05, log-rank-test).

CONCLUSION

The iPD score provides oncologists with a new evaluation, computable at first-progression, to decide if treatment should be continued (for the GP group), or immediately changed for the PP and CP groups. Further validation on larger cohorts is needed to prove its efficacy in clinical practice.

Predicting immunotherapy outcomes in patients with MSI tumors using NLR and CT global tumor volume

BACKGROUND

Anti-PD-(L)1 treatment is indicated for patients with mismatch repair-deficient (MMRD) tumors, regardless of tumor origin. However, the response rate is highly heterogeneous across MMRD tumors. The objective of the study is to find a score that predicts anti-PD-(L)1 response in patients with MMRD tumors.

METHODS

Sixty-one patients with various origin of MMRD tumors and treated with anti-PD-(L)1 were retrospectively included in this study. An expert radiologist annotated all tumors present at the baseline and first evaluation CT-scans for all the patients by circumscribing them on their largest axial axis (single slice), allowing us to compute an approximation of their tumor volume. In total, 2120 lesions were annotated, which led to the computation of the total tumor volume for each patient. The RECIST sum of target lesions' diameters and neutrophile-to-lymphocyte (NLR) were also reported at both examinations. These parameters were determined at baseline and first evaluation and the variation between the first evaluation and baseline was calculated, to determine a comprehensive score for overall survival (OS) and progression-free survival (PFS).

RESULTS

Total tumor volume at baseline was found to be significantly correlated to the OS (p-value: 0.005) and to the PFS (p-value:<0.001). The variation of the RECIST sum of target lesions' diameters, total tumor volume and NLR were found to be significantly associated to the OS (p-values:<0.001, 0.006,<0.001 respectively) and to the PFS (<0.001,<0.001, 0.007 respectively). The concordance score combining total tumor volume and NLR variation was better at stratifying patients compared to the tumor volume or NLR taken individually according to the OS (pairwise log-rank test p-values: 0.033,<0.001, 0.002) and PFS (pairwise log-rank test p-values: 0.041,<0.001, 0.003).

CONCLUSION

Total tumor volume appears to be a prognostic biomarker of anti-PD-(L)1 response to immunotherapy in metastatic patients with MMRD tumors. Combining tumor volume and NLR with a simple concordance score stratifies patients well according to their survival and offers a good predictive measure of response to immunotherapy.

Machine learning predicts response to TNF inhibitors in rheumatoid arthritis: results on the ESPOIR and ABIRISK cohorts

Objectives

Around 30% of patients with rheumatoid arthritis (RA) do not respond to tumour necrosis factor inhibitors (TNFi). We aimed to predict patient response to TNFi using machine learning on simple clinical and biological data.

Methods

We used data from the RA ESPOIR cohort to train our models. The endpoints were the EULAR response and the change in Disease Activity Score (DAS28). We compared the performances of multiple models (linear regression, random forest, XGBoost and CatBoost) on the training set and cross-validated them using the area under the receiver operating characteristic curve (AUROC) or the mean squared error. The best model was then evaluated on a replication cohort (ABIRISK).

Results

We included 161 patients from ESPOIR and 118 patients from ABIRISK. The key selected features were DAS28, lymphocytes, ALT (aspartate aminotransferase), neutrophils, age, weight, and smoking status. When predicting EULAR response, CatBoost achieved the best performances of the four tested models. It reached an AUROC of 0.72 (0.68–0.73) on the train set (ESPOIR). Better results were obtained on the train set when etanercept and monoclonal antibodies were analysed separately. On the test set (ABIRISK), these models respectively achieved on AUROC of 0.70 (0.57–0.82) and 0.71 (0.55–0.86). Two decision thresholds were tested. The first prioritised a high confidence in identifying responders and yielded a confidence up to 90% for predicting response. The second prioritised a high confidence in identifying inadequate responders and yielded a confidence up to 70% for predicting non-response. The change in DAS28 was predicted with an average error of 1.1 DAS28 points.

Conclusion

The machine learning models developed allowed predicting patient response to TNFi exclusively using data available in clinical routine.

Abstract P4-07-17: Association between plasma-based sequential windowed acquisition mass spectrometry (SWATH-MS) and invasive disease free survival (iDFS) in HR+/HER2- early breast cancer in the CANTO cohort

Background: The definition of breast cancer (BC) prognosis has historically relied on clinico-pathological factors. Novel omics markers including proteomic analyses could improve our understanding of the biological host drivers of breast cancer recurrence and survival. We aimed at identifying patients (pts) at high risk of recurrence based on proteomic markers in plasma.Methods: CANTO is a multicenter, prospective cohort study of stage I-III BCS (NCT01993498). Plasma samples were collected on HR+/HER2- pts at diagnosis (dx) and analyzed by SWATH-MS, implemented by Biognosys AG (Schlieren, Switzerland), resulting in a relative quantification of the abundance of 500 proteins in the plasma. A Cox model was fitted to estimate to associate proteomic and clinical variables with the primary endpoint IDFS Clinical covariates consisted of age, stage and grade. An adaptive Lasso method was used to perform model selection. The discrimination performances of the model were assessed on 100 random train-test partitions of the cohort. Results: There were 457 pts with analyzed plasma samples. The median age at dx was 59.3 years, and the repartition of cancer stage was 52% for stage I, 37% for stage II and 11% for stage III. The mean duration of follow-up was 5.4 years, and 53 (11.5%) IDFS events (non local recurrences, second primary cancers and deaths) were reported. In total, 7 proteins were selected by the adaptive Lasso process; associated with the age, the stage and the grade at dx, 3 proteins were retained as having a significant impact on the IDFS: GTP-binding nuclear protein Ran (RAN), involved in cell division and GTP metabolic process, C4b-binding protein alpha-chain (C4BPA), involved in complement activation, and prothrombin (THRB), involved in acute-phase response and blood activation. Concordance indices were computed on 100 random test subsets of the cohort for the model with clinical variables only (0.67+/- 0.08), for the model with selected protein features only (0.74 +/- 0.07) and for the model with both proteomic and clinical covariates (0.75 +/-0.06). Conclusion: The discrimination performances of the estimated model suggest that proteomics provide relevant markers associated with BC prognosis. Validation on an independent validation set is required. Host related plasma proteins represent an avenue worth exploring to improve our understanding of BC relapse risk

Table 1.Estimated hazard ratios of the linear Cox model.FeaturesHR* (95% CI)p-valuesRAN (for 1 SD increase)0.66 (0.51-0.85)&lt;0.005THRB (for 1 SD increase)1.43 (0.99-2.06)0.05C4BPA (for 1 SD increase)1.44 (1.02-2.02)0.04stage--II vs I1.68 (0.82-3.46)0.16III vs I4.29 (1.88-9.75)&lt;0.005HR = hazard ratio CI = confidence interval * adjusted by age and grade

Citation Format: Antonin Della Noce, Stergios Christodoulidis, Antonio Di Meglio, Julie Havas, Alicia Tran-Dien, Fabrice André, Ines Vaz-Luis, Paul-Henry Cournède, Stefan Michiels. Association between plasma-based sequential windowed acquisition mass spectrometry (SWATH-MS) and invasive disease free survival (iDFS) in HR+/HER2- early breast cancer in the CANTO cohort [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P4-07-17.

Comprehensive landscape of immune-checkpoints uncovered in clear cell renal cell carcinoma reveals new and emerging therapeutic targets

Clear cell renal cell carcinoma (ccRCC) constitutes the most common renal cell carcinoma subtype and has long been recognized as an immunogenic cancer. As such, significant attention has been directed toward optimizing immune-checkpoints (IC)-based therapies. Despite proven benefits, a substantial number of patients remain unresponsive to treatment, suggesting that yet unreported, immunosuppressive mechanisms coexist within tumors and their microenvironment. Here, we comprehensively analyzed and ranked forty-four immune-checkpoints expressed in ccRCC on the basis of in-depth analysis of RNAseq data collected from the TCGA database and advanced statistical methods designed to obtain the group of checkpoints that best discriminates tumor from healthy tissues. Immunohistochemistry and flow cytometry confirmed and enlarged the bioinformatics results. In particular, by using the recursive feature elimination method, we show that HLA-G, B7H3, PDL-1 and ILT2 are the most relevant genes that characterize ccRCC. Notably, ILT2 expression was detected for the first time on tumor cells. The levels of other ligand-receptor pairs such as CD70:CD27; 4-1BB:4-1BBL; CD40:CD40L; CD86:CTLA4; MHC-II:Lag3; CD200:CD200R; CD244:CD48 were also found highly expressed in tumors compared to adjacent non-tumor tissues. Collectively, our approach provides a comprehensible classification of forty-four IC expressed in ccRCC, some of which were never reported before to be co-expressed in ccRCC. In addition, the algorithms used allowed identifying the most relevant group that best discriminates tumor from healthy tissues. The data can potentially assist on the choice of valuable immune-therapy targets which hold potential for the development of more effective anti-tumor treatments.

A new methodology based on sensitivity analysis to simplify the recalibration of functional-structural plant models in new conditions

Background and Aims

Functional-structural plant models (FSPMs) describe explicitly the interactions between plants and their environment at organ to plant scale. However, the high level of description of the structure or model mechanisms makes this type of model very complex and hard to calibrate. A two-step methodology to facilitate the calibration process is proposed here.

Methods

First, a global sensitivity analysis method was applied to the calibration loss function. It provided first-order and total-order sensitivity indexes that allow parameters to be ranked by importance in order to select the most influential ones. Second, the Akaike information criterion (AIC) was used to quantify the model's quality of fit after calibration with different combinations of selected parameters. The model with the lowest AIC gives the best combination of parameters to select. This methodology was validated by calibrating the model on an independent data set (same cultivar, another year) with the parameters selected in the second step. All the parameters were set to their nominal value; only the most influential ones were re-estimated.

Key Results

Sensitivity analysis applied to the calibration loss function is a relevant method to underline the most significant parameters in the estimation process. For the studied winter oilseed rape model, 11 out of 26 estimated parameters were selected. Then, the model could be recalibrated for a different data set by re-estimating only three parameters selected with the model selection method.

Conclusions

Fitting only a small number of parameters dramatically increases the efficiency of recalibration, increases the robustness of the model and helps identify the principal sources of variation in varying environmental conditions. This innovative method still needs to be more widely validated but already gives interesting avenues to improve the calibration of FSPMs.

Leaf Segmentation and Tracking in Arabidopsis thaliana Combined to an Organ-Scale Plant Model for Genotypic Differentiation

A promising method for characterizing the phenotype of a plant as an interaction between its genotype and its environment is to use refined organ-scale plant growth models that use the observation of architectural traits, such as leaf area, containing a lot of information on the whole history of the functioning of the plant. The Phenoscope, a high-throughput automated platform, allowed the acquisition of zenithal images of Arabidopsis thaliana over twenty one days for 4 different genotypes. A novel image processing algorithm involving both segmentation and tracking of the plant leaves allows to extract areas of the latter. First, all the images in the series are segmented independently using a watershed-based approach. A second step based on ellipsoid-shaped leaves is then applied on the segments found to refine the segmentation. Taking into account all the segments at every time, the whole history of each leaf is reconstructed by choosing recursively through time the most probable segment achieving the best score, computed using some characteristics of the segment such as its orientation, its distance to the plant mass center and its area. These results are compared to manually extracted segments, showing a very good accordance in leaf rank and that they therefore provide low-biased data in large quantity for leaf areas. Such data can therefore be exploited to design an organ-scale plant model adapted from the existing GreenLab model for A. thaliana and subsequently parameterize it. This calibration of the model parameters should pave the way for differentiation between the Arabidopsis genotypes.

Modeling tree crown dynamics with 3D partial differential equations

We characterize a tree's spatial foliage distribution by the local leaf area density. Considering this spatially continuous variable allows to describe the spatiotemporal evolution of the tree crown by means of 3D partial differential equations. These offer a framework to rigorously take locally and adaptively acting effects into account, notably the growth toward light. Biomass production through photosynthesis and the allocation to foliage and wood are readily included in this model framework. The system of equations stands out due to its inherent dynamic property of self-organization and spontaneous adaptation, generating complex behavior from even only a few parameters. The density-based approach yields spatially structured tree crowns without relying on detailed geometry. We present the methodological fundamentals of such a modeling approach and discuss further prospects and applications.

Modeling spatial competition for light in plant populations with the porous medium equation

We consider a plant's local leaf area index as a spatially continuous variable, subject to particular reaction-diffusion dynamics of allocation, senescence and spatial propagation. The latter notably incorporates the plant's tendency to form new leaves in bright rather than shaded locations. Applying a generalized Beer-Lambert law allows to link existing foliage to production dynamics. The approach allows for inter-individual variability and competition for light while maintaining robustness-a key weakness of comparable existing models. The analysis of the single plant case leads to a significant simplification of the system's key equation when transforming it into the well studied porous medium equation. Confronting the theoretical model to experimental data of sugar beet populations, differing in configuration density, demonstrates its accuracy.

NEMA, a functional-structural model of nitrogen economy within wheat culms after flowering. II. Evaluation and sensitivity analysis

BACKGROUND AND AIMS

Simulating nitrogen economy in crop plants requires formalizing the interactions between soil nitrogen availability, root nitrogen acquisition, distribution between vegetative organs and remobilization towards grains. This study evaluates and analyses the functional-structural and mechanistic model of nitrogen economy, NEMA (Nitrogen Economy Model within plant Architecture), developed for winter wheat (Triticum aestivum) after flowering.

METHODS

NEMA was calibrated for field plants under three nitrogen fertilization treatments at flowering. Model behaviour was investigated and sensitivity to parameter values was analysed.

KEY RESULTS

Nitrogen content of all photosynthetic organs and in particular nitrogen vertical distribution along the stem and remobilization patterns in response to fertilization were simulated accurately by the model, from Rubisco turnover modulated by light intercepted by the organ and a mobile nitrogen pool. This pool proved to be a reliable indicator of plant nitrogen status, allowing efficient regulation of nitrogen acquisition by roots, remobilization from vegetative organs and accumulation in grains in response to nitrogen treatments. In our simulations, root capacity to import carbon, rather than carbon availability, limited nitrogen acquisition and ultimately nitrogen accumulation in grains, while Rubisco turnover intensity mostly affected dry matter accumulation in grains.

CONCLUSIONS

NEMA enabled interpretation of several key patterns usually observed in field conditions and the identification of plausible processes limiting for grain yield, protein content and root nitrogen acquisition that could be targets for plant breeding; however, further understanding requires more mechanistic formalization of carbon metabolism. Its strong physiological basis and its realistic behaviour support its use to gain insights into nitrogen economy after flowering.

NEMA, a functional-structural model of nitrogen economy within wheat culms after flowering. I. Model description

BACKGROUND AND AIMS

Models simulating nitrogen use by plants are potentially efficient tools to optimize the use of fertilizers in agriculture. Most crop models assume that a target nitrogen concentration can be defined for plant tissues and formalize a demand for nitrogen, depending on the difference between the target and actual nitrogen concentrations. However, the teleonomic nature of the approach has been criticized. This paper proposes a mechanistic model of nitrogen economy, NEMA (Nitrogen Economy Model within plant Architecture), which links nitrogen fluxes to nitrogen concentration and physiological processes.

METHODS

A functional-structural approach is used: plant aerial parts are described in a botanically realistic way and physiological processes are expressed at the scale of each aerial organ or root compartment as a function of local conditions (light and resources).

KEY RESULTS

NEMA was developed for winter wheat (Triticum aestivum) after flowering. The model simulates the nitrogen (N) content of each photosynthetic organ as regulated by Rubisco turnover, which depends on intercepted light and a mobile N pool shared by all organs. This pool is enriched by N acquisition from the soil and N release from vegetative organs, and is depleted by grain uptake and protein synthesis in vegetative organs; NEMA accounts for the negative feedback from circulating N on N acquisition from the soil, which is supposed to follow the activities of nitrate transport systems. Organ N content and intercepted light determine dry matter production via photosynthesis, which is distributed between organs according to a demand-driven approach.

CONCLUSIONS

NEMA integrates the main feedbacks known to regulate plant N economy. Other novel features are the simulation of N for all photosynthetic tissues and the use of an explicit description of the plant that allows how the local environment of tissues regulates their N content to be taken into account. We believe this represents an appropriate frame for modelling nitrogen in functional-structural plant models. A companion paper will present model evaluation and analysis.

Characterization of the interactions between architecture and source-sink relationships in winter oilseed rape (Brassica napus) using the GreenLab model

BACKGROUND AND AIMS

This study aimed to characterize the interaction between architecture and source-sink relationships in winter oilseed rape (WOSR): do the costs of ramification compromise the source-sink ratio during seed filling? The GreenLab model is a good candidate to address this question because it has been already used to describe interactions between source-sink relationships and architecture for other species. However, its adaptation to WOSR is a challenge because of the complexity of its developmental scheme, especially during the reproductive phase.

METHODS

Equations were added in GreenLab to compute expansion delays for ramification, flowering of each axis and photosynthesis of pods including the energetic cost of oil synthesis. Experimental field data were used to estimate morphological parameters while source-sink parameters of the model were estimated by adjustment of model outputs to the data. Ecophysiological outputs were used to assess the sources/sink relationships during the whole growth cycle.

KEY RESULTS

First results indicated that, at the plant scale, the model correctly simulates the dynamics of organ growth. However, at the organ scale, errors were observed that could be explained either by secondary growth that was not incorporated or by uncertainties in morphological parameters (durations of expansion and life). Ecophysiological outputs highlighted the dramatic negative impact of ramification on the source-sink ratio, as well as the decrease in this ratio during seed filling despite pod envelope photosynthesis that allowed significant biomass production to be maintained.

CONCLUSIONS

This work is a promising first step in the construction of a structure-function model for a plant as complex as WOSR. Once tested for other environments and/or genotypes, the model can be used for studies on WOSR architectural plasticity.

A dynamic model of plant growth with interactions between development and functional mechanisms to study plant structural plasticity related to trophic competition

BACKGROUND AND AIMS

The strong influence of environment and functioning on plant organogenesis has been well documented by botanists but is poorly reproduced in most functional-structural models. In this context, a model of interactions is proposed between plant organogenesis and plant functional mechanisms.

METHODS

The GreenLab model derived from AMAP models was used. Organogenetic rules give the plant architecture, which defines an interconnected network of organs. The plant is considered as a collection of interacting 'sinks' that compete for the allocation of photosynthates coming from 'sources'. A single variable characteristic of the balance between sources and sinks during plant growth controls different events in plant development, such as the number of branches or the fruit load.

KEY RESULTS

Variations in the environmental parameters related to light and density induce changes in plant morphogenesis. Architecture appears as the dynamic result of this balance, and plant plasticity expresses itself very simply at different levels: appearance of branches and reiteration, number of organs, fructification and adaptation of ecophysiological characteristics.

CONCLUSIONS

The modelling framework serves as a tool for theoretical botany to explore the emergence of specific morphological and architectural patterns and can help to understand plant phenotypic plasticity and its strategy in response to environmental changes.

Using a mathematical model to evaluate the trophic and non-trophic determinants of axis development in grapevine

The grapevine (Vitis vinifera L.) shoot is a complex modular branching system, with one primary axis and many secondary axes organised into a repetitive structure of three successive phytomers (P0-P1-P2). P1-P2 phytomers bear one tendril or cluster, whereas P0 phytomers bear no tendrils or clusters. Axis development displays a high variability, due, partly, to trophic competition. The aim of this study was to estimate changes in trophic competition within the shoot, and to relate plasticity in axis development to changes in trophic competition. 'Grenache N.' and 'Syrah' cultivars were grown with two contrasting levels of cluster load. Organogenesis and organ mass were measured during shoot development. Changes in trophic competition were estimated, using the solver functions of the GreenLab model. Internodes and clusters were strong sinks. They affected the shoot development to the same extent, but the internodes showed an earlier effect. The cessation of development of the secondary axis was affected by trophic competition, but the primary axis continued to develop, regardless of trophic competition. Secondary axes differed in sensitivity to trophic competition as a function of two criteria: their type and their size. The most highly developed axes were less affected than the smaller axes, and secondary axes arising from a P0 phytomer were also less affected than secondary axes arising from a P1 or P2 phytomer.

A model-based analysis of the dynamics of carbon balance at the whole-plant level in Arabidopsis thaliana

Arabidopsis thaliana (L.) Heynh. is used as a model plant in many research projects. However, few models simulate its growth at the whole-plant scale. The present study describes the first model of Arabidopsis growth integrating organogenesis, morphogenesis and carbon-partitioning processes for aerial and subterranean parts of the plant throughout its development. The objective was to analyse competition among sinks as they emerge from patterns of plant structural development. The model was adapted from the GreenLab model and was used to estimate organ sink strengths by optimisation against biomass measurements. Dry biomass production was calculated by a radiation use efficiency-based approach. Organogenesis processes were parameterised based on experimental data. The potential of this model for growth analysis was assessed using the Columbia ecotype, which was grown in standard environmental conditions. Three phases were observed in the overall time course of trophic competition within the plant. In the vegetative phase, no competition was observed. In the reproductive phase, competition increased with a strong increase when lateral inflorescences developed. Roots and internodes and structures bearing siliques were strong sinks and had a similar impact on competition. The application of the GreenLab model to the growth analysis of A. thaliana provides new insights into source-sink relationships as functions of phenology and morphogenesis.

Parametric identification of a functional-structural tree growth model and application to beech trees (Fagus sylvatica)

Functional-structural models provide detailed representations of tree growth and their application to forestry seems full of prospects. However, owing to the complexity of tree architecture, parametric identification of such models remains a critical issue. We present the GreenLab approach for modelling tree growth. It simulates tree growth plasticity in response to changes of their internal level of trophic competition, especially topological development and cambial growth. The model includes a simplified representation of tree architecture, based on a species-specific description of branching patterns. We study whether those simplifications allow enough flexibility to reproduce with the same set of parameters the growth of two observed understorey beech trees (Fagus sylvatica L.) of different ages in different environmental conditions. The parametric identification of the model is global, i.e. all parameters are estimated simultaneously, potentially providing a better description of interactions between sub-processes. As a result, the source-sink dynamics throughout tree development is retrieved. Simulated and measured trees were compared for their trunk profiles (fresh masses and dimensions of every growth units, ring diameters at different heights) and compartment masses of their order 2 branches. Possible improvements of this method by including topological criteria are discussed.

Computing competition for light in the GREENLAB model of plant growth: a contribution to the study of the effects of density on resource acquisition and architectural development

BACKGROUND AND AIMS

The dynamical system of plant growth GREENLAB was originally developed for individual plants, without explicitly taking into account interplant competition for light. Inspired by the competition models developed in the context of forest science for mono-specific stands, we propose to adapt the method of crown projection onto the x-y plane to GREENLAB, in order to study the effects of density on resource acquisition and on architectural development.

METHODS

The empirical production equation of GREENLAB is extrapolated to stands by computing the exposed photosynthetic foliage area of each plant. The computation is based on the combination of Poisson models of leaf distribution for all the neighbouring plants whose crown projection surfaces overlap. To study the effects of density on architectural development, we link the proposed competition model to the model of interaction between functional growth and structural development introduced by Mathieu (2006, PhD Thesis, Ecole Centrale de Paris, France).

KEY RESULTS AND CONCLUSIONS

The model is applied to mono-specific field crops and forest stands. For high-density crops at full cover, the model is shown to be equivalent to the classical equation of field crop production (Howell and Musick, 1985, in Les besoins en eau des cultures; Paris: INRA Editions). However, our method is more accurate at the early stages of growth (before cover) or in the case of intermediate densities. It may potentially account for local effects, such as uneven spacing, variation in the time of plant emergence or variation in seed biomass. The application of the model to trees illustrates the expression of plant plasticity in response to competition for light. Density strongly impacts on tree architectural development through interactions with the source-sink balances during growth. The effects of density on tree height and radial growth that are commonly observed in real stands appear as emerging properties of the model.

Parameter optimization and field validation of the functional-structural model GREENLAB for maize at different population densities

BACKGROUND AND AIMS

Plant population density (PPD) influences plant growth greatly. Functional-structural plant models such as GREENLAB can be used to simulate plant development and growth and PPD effects on plant functioning and architectural behaviour can be investigated. This study aims to evaluate the ability of GREENLAB to predict maize growth and development at different PPDs.

METHODS

Two field experiments were conducted on irrigated fields in the North China Plain with a block design of four replications. Each experiment included three PPDs: 2.8, 5.6 and 11.1 plants m(-2). Detailed observations were made on the dimensions and fresh biomass of above-ground plant organs for each phytomer throughout the seasons. Growth stage-specific target files (a description of plant organ weight and dimension according to plant topological structure) were established from the measured data required for GREENLAB parameterization. Parameter optimization was conducted using a generalized least square method for the entire growth cycles for all PPDs and years. Data from in situ plant digitization were used to establish geometrical symbol files for organs that were then applied to translate model output directly into 3-D representation for each time step of the model execution.

KEY RESULTS

The analysis indicated that the parameter values of organ sink variation function, and the values of most of the relative sink strength parameters varied little among years and PPDs, but the biomass production parameter, computed plant projection surface and internode relative sink strength varied with PPD. Simulations of maize plant growth based on the fitted parameters were reasonably good as indicated by the linearity and slopes similar to unity for the comparison of simulated and observed values. Based on the parameter values fitted from different PPDs, shoot (including vegetative and reproductive parts of the plant) and cob fresh biomass for other PPDs were simulated. Three-dimensional representation of individual plant and plant stand from the model output with two contrasting PPDs were presented with which the PPD effect on plant growth can be easily recognized.

CONCLUSIONS

This study showed that GREENLAB model has the ability to capture plant plasticity induced by PPD. The relatively stable parameter values strengthened the hypothesis that one set of equations can govern dynamic organ growth. With further validation, this model can be used for agronomic applications such as yield optimization.

Rhythms and alternating patterns in plants as emergent properties of a model of interaction between development and functioning

BACKGROUND AND AIMS

To model plasticity of plants in their environment, a new version of the functional-structural model GREENLAB has been developed with full interactions between architecture and functioning. Emergent properties of this model were revealed by simulations, in particular the automatic generation of rhythms in plant development. Such behaviour can be observed in natural phenomena such as the appearance of fruit (cucumber or capsicum plants, for example) or branch formation in trees.

METHODS

In the model, a single variable, the source-sink ratio controls different events in plant architecture. In particular, the number of fruits and branch formation are determined as increasing functions of this ratio. For some sets of well-chosen parameters of the model, the dynamical evolution of the ratio during plant growth generates rhythms.

KEY RESULTS AND CONCLUSIONS

Cyclic patterns in branch formation or fruit appearance emerge without being forced by the model. The model is based on the theory of discrete dynamical systems. The mathematical formalism helps us to explain rhythm generation and to control the behaviour of the system. Rhythms can appear during both the exponential and stabilized phases of growth, but the causes are different as shown by an analytical study of the system. Simulated plant behaviours are very close to those observed on real plants. With a small number of parameters, the model gives very interesting results from a qualitative point of view. It will soon be subjected to experimental data to estimate the model parameters.

Quantitative genetics and functional-structural plant growth models: simulation of quantitative trait loci detection for model parameters and application to potential yield optimization

BACKGROUND AND AIMS

Prediction of phenotypic traits from new genotypes under untested environmental conditions is crucial to build simulations of breeding strategies to improve target traits. Although the plant response to environmental stresses is characterized by both architectural and functional plasticity, recent attempts to integrate biological knowledge into genetics models have mainly concerned specific physiological processes or crop models without architecture, and thus may prove limited when studying genotype x environment interactions. Consequently, this paper presents a simulation study introducing genetics into a functional-structural growth model, which gives access to more fundamental traits for quantitative trait loci (QTL) detection and thus to promising tools for yield optimization.

METHODS

The GREENLAB model was selected as a reasonable choice to link growth model parameters to QTL. Virtual genes and virtual chromosomes were defined to build a simple genetic model that drove the settings of the species-specific parameters of the model. The QTL Cartographer software was used to study QTL detection of simulated plant traits. A genetic algorithm was implemented to define the ideotype for yield maximization based on the model parameters and the associated allelic combination.

KEY RESULTS AND CONCLUSIONS

By keeping the environmental factors constant and using a virtual population with a large number of individuals generated by a Mendelian genetic model, results for an ideal case could be simulated. Virtual QTL detection was compared in the case of phenotypic traits--such as cob weight--and when traits were model parameters, and was found to be more accurate in the latter case. The practical interest of this approach is illustrated by calculating the parameters (and the corresponding genotype) associated with yield optimization of a GREENLAB maize model. The paper discusses the potentials of GREENLAB to represent environment x genotype interactions, in particular through its main state variable, the ratio of biomass supply over demand.