<p>(A) shows the representation of a rule in the multi-dimensional variable space (left), p... more <p>(A) shows the representation of a rule in the multi-dimensional variable space (left), projected on each variable (middle) and as binary variable spanning over all the subjects (right). (B) shows the selection of the candidate rules proceeding in the following two steps: thresholding on two rule quality measures (left) and a minimization procedure to remove redundant rules (right).</p
<p>This figure shows the frequency (y axis) of each bucket selected over the 100 global mod... more <p>This figure shows the frequency (y axis) of each bucket selected over the 100 global models built on the metabolomic data (x axis). The red horizontal line corresponds to a frequency threshold of 50. Buckets with a frequency higher than this threshold are labeled with their corresponding metabolites.</p
Multivariate classification methods using explanatory and predictive models are necessary for cha... more Multivariate classification methods using explanatory and predictive models are necessary for characterizing subgroups of patients according to their risk profiles. Popular methods include logistic regression and classification trees with performances that vary according to the nature and the characteristics of the dataset. In the context of imported malaria, we aimed at classifying severity criteria based on a heterogeneous patient population. We investigated these approaches by implementing two different strategies: L1 logistic regression (L1LR) that models a single global solution and classification trees that model multiple local solutions corresponding to discriminant subregions of the feature space. For each strategy, we built a standard model, and a sparser version of it. As an alternative to pruning, we explore a promising approach that first constrains the tree model with an L1LR-based feature selection, an approach we called L1LR-Tree. The objective is to decrease its vuln...
<b>Copyright information:</b>Taken from "Incremental and unifying modelling form... more <b>Copyright information:</b>Taken from "Incremental and unifying modelling formalism for biological interaction networks"http://www.biomedcentral.com/1471-2105/8/433BMC Bioinformatics 2007;8():433-433.Published online 8 Nov 2007PMCID:PMC2200675.
<b>Copyright information:</b>Taken from "Incremental and unifying modelling form... more <b>Copyright information:</b>Taken from "Incremental and unifying modelling formalism for biological interaction networks"http://www.biomedcentral.com/1471-2105/8/433BMC Bioinformatics 2007;8():433-433.Published online 8 Nov 2007PMCID:PMC2200675. Functions and come, on one hand, from the MIN topology and the information on the stoichiometry of the reaction, and on the other hand, from the reaction attribute. At this stage, the coherence of both informations should be checked by an expert. In these equations and have a definite signature reflecting the impact of the catalyzers and inhibitors on the reactions.
subgroup discovery algorithm Margaux Luck, Gildas Bertho, Eric Thervet, Philippe Beaune, François... more subgroup discovery algorithm Margaux Luck, Gildas Bertho, Eric Thervet, Philippe Beaune, François d’Ormesson, Mathilde Bateson, Anastasia Yartseva, Cécilia Damon, Nicolas Pallet 1 Hypercube Institute, Paris, France; 2 Paris Descartes University, Paris, France; 3 CNRS UMR 8601, Paris, France; 4 Magnetic Nuclear Resonance Plateform “Metabo Paris-Santé”, CICB-Paris, Paris, France; 5 Renal Division, Georges Pompidou European Hospital, Assistance Publique Hôpitaux de Paris, Paris, France; 6 INSERM U1147, Paris, France; 7 Clinical Chemistry Department, Georges Pompidou European Hospital, Paris, France H Nuclear Magnetic Resonance (NMR) spectroscopy is widely used for the identification of metabolites in biofluids and enables the understanding of the underlying biological mechanisms of a disease. Within the study of chronic kidney disease (CKD), the analysis of metabolomic variations, acquired with H NMR in urine, may be useful for the diagnosis of the different CKD severity stages. Due to...
The actual criteria for the classification of the different forms of imported malaria are complex... more The actual criteria for the classification of the different forms of imported malaria are complex and do not take into account the heterogeneity of the individual profiles. Multivariate classification methods using explanatory and predictive models are necessary to characterize groups with a high risk of developing severe forms of imported malaria. We investigate two standard approaches implementing two different strategies: L1 logistic regression that models a single global solution, which is a linear combination of a subset of the input features, and classification trees that models multiple local solutions corresponding to discriminate sub regions of the feature space. As an alternative to pruning, which limits the complexity of the decision tree by removing unstable branches once the model is built, we explore an original approach known as L1 LR-Tree, which combines the two previous strategies. This combined method constrains the dimension of the initial set of features before f...
Multivariate classification methods using explanatory and predictive models are necessary for cha... more Multivariate classification methods using explanatory and predictive models are necessary for characterizing subgroups of patients according to their risk profiles. Popular methods include logistic regression and classification trees with performances that vary according to the nature and the characteristics of the dataset. In the context of imported malaria, we aimed at classifying severity criteria based on a heterogeneous patient population. We investigated these approaches by implementing two different strategies: L1 logistic regression (L1LR) that models a single global solution and classification trees that model multiple local solutions corresponding to discriminant subregions of the feature space. For each strategy, we built a standard model, and a sparser version of it. As an alternative to pruning, we explore a promising approach that first constrains the tree model with an L1LR-based feature selection, an approach we called L1LR-Tree. The objective is to decrease its vuln...
... Views on the choice of simulation tools, compared in the context of the same model: the bacte... more ... Views on the choice of simulation tools, compared in the context of the same model: the bacteriophage Lambda. Anastasia Yartseva 1 , Denis Mestivier 2 , Pierre-Yves Boëlle 3 , Adrien Richard 1 , Guillaume Hutzler 1. (2005). ...
Metabolic profiling, the study of changes in the concentration of the metabolites in the organism... more Metabolic profiling, the study of changes in the concentration of the metabolites in the organism induced by biological differences within subpopulations, has to deal with a very large amount of complex data. It therefore requires the use of powerful data processing and machine learning methods. To overcome over-fitting, a common concern in metabolic profiling where the number of features is often much larger than the number of observations, many predictive analyses combined dimension reduction techniques with multivariate predictive linear modeling. Moreover, they built a global model that identifies biomarkers predictive of the output of interest giving their overall trend variations. However, this fails to capture local biological phenomena underlying subgroups of subjects. More recently, local exploration methods based on decision trees approaches have been applied in metabolomics but they only explore random parts of the feature space. In this study, we used a supervised rule-m...
... Ateliers de démonstration et de comparaison d'outils de modélisation autour de modèles b... more ... Ateliers de démonstration et de comparaison d'outils de modélisation autour de modèles biologiques communs. Anastasia Yartseva 1 , Denis Mestivier 2 , Pierre-Yves Boëlle 3 , Adrien Richard 1 , Guillaume Hutzler 1. (2005). ...
<p>(A) shows the representation of a rule in the multi-dimensional variable space (left), p... more <p>(A) shows the representation of a rule in the multi-dimensional variable space (left), projected on each variable (middle) and as binary variable spanning over all the subjects (right). (B) shows the selection of the candidate rules proceeding in the following two steps: thresholding on two rule quality measures (left) and a minimization procedure to remove redundant rules (right).</p
<p>This figure shows the frequency (y axis) of each bucket selected over the 100 global mod... more <p>This figure shows the frequency (y axis) of each bucket selected over the 100 global models built on the metabolomic data (x axis). The red horizontal line corresponds to a frequency threshold of 50. Buckets with a frequency higher than this threshold are labeled with their corresponding metabolites.</p
Multivariate classification methods using explanatory and predictive models are necessary for cha... more Multivariate classification methods using explanatory and predictive models are necessary for characterizing subgroups of patients according to their risk profiles. Popular methods include logistic regression and classification trees with performances that vary according to the nature and the characteristics of the dataset. In the context of imported malaria, we aimed at classifying severity criteria based on a heterogeneous patient population. We investigated these approaches by implementing two different strategies: L1 logistic regression (L1LR) that models a single global solution and classification trees that model multiple local solutions corresponding to discriminant subregions of the feature space. For each strategy, we built a standard model, and a sparser version of it. As an alternative to pruning, we explore a promising approach that first constrains the tree model with an L1LR-based feature selection, an approach we called L1LR-Tree. The objective is to decrease its vuln...
<b>Copyright information:</b>Taken from "Incremental and unifying modelling form... more <b>Copyright information:</b>Taken from "Incremental and unifying modelling formalism for biological interaction networks"http://www.biomedcentral.com/1471-2105/8/433BMC Bioinformatics 2007;8():433-433.Published online 8 Nov 2007PMCID:PMC2200675.
<b>Copyright information:</b>Taken from "Incremental and unifying modelling form... more <b>Copyright information:</b>Taken from "Incremental and unifying modelling formalism for biological interaction networks"http://www.biomedcentral.com/1471-2105/8/433BMC Bioinformatics 2007;8():433-433.Published online 8 Nov 2007PMCID:PMC2200675. Functions and come, on one hand, from the MIN topology and the information on the stoichiometry of the reaction, and on the other hand, from the reaction attribute. At this stage, the coherence of both informations should be checked by an expert. In these equations and have a definite signature reflecting the impact of the catalyzers and inhibitors on the reactions.
subgroup discovery algorithm Margaux Luck, Gildas Bertho, Eric Thervet, Philippe Beaune, François... more subgroup discovery algorithm Margaux Luck, Gildas Bertho, Eric Thervet, Philippe Beaune, François d’Ormesson, Mathilde Bateson, Anastasia Yartseva, Cécilia Damon, Nicolas Pallet 1 Hypercube Institute, Paris, France; 2 Paris Descartes University, Paris, France; 3 CNRS UMR 8601, Paris, France; 4 Magnetic Nuclear Resonance Plateform “Metabo Paris-Santé”, CICB-Paris, Paris, France; 5 Renal Division, Georges Pompidou European Hospital, Assistance Publique Hôpitaux de Paris, Paris, France; 6 INSERM U1147, Paris, France; 7 Clinical Chemistry Department, Georges Pompidou European Hospital, Paris, France H Nuclear Magnetic Resonance (NMR) spectroscopy is widely used for the identification of metabolites in biofluids and enables the understanding of the underlying biological mechanisms of a disease. Within the study of chronic kidney disease (CKD), the analysis of metabolomic variations, acquired with H NMR in urine, may be useful for the diagnosis of the different CKD severity stages. Due to...
The actual criteria for the classification of the different forms of imported malaria are complex... more The actual criteria for the classification of the different forms of imported malaria are complex and do not take into account the heterogeneity of the individual profiles. Multivariate classification methods using explanatory and predictive models are necessary to characterize groups with a high risk of developing severe forms of imported malaria. We investigate two standard approaches implementing two different strategies: L1 logistic regression that models a single global solution, which is a linear combination of a subset of the input features, and classification trees that models multiple local solutions corresponding to discriminate sub regions of the feature space. As an alternative to pruning, which limits the complexity of the decision tree by removing unstable branches once the model is built, we explore an original approach known as L1 LR-Tree, which combines the two previous strategies. This combined method constrains the dimension of the initial set of features before f...
Multivariate classification methods using explanatory and predictive models are necessary for cha... more Multivariate classification methods using explanatory and predictive models are necessary for characterizing subgroups of patients according to their risk profiles. Popular methods include logistic regression and classification trees with performances that vary according to the nature and the characteristics of the dataset. In the context of imported malaria, we aimed at classifying severity criteria based on a heterogeneous patient population. We investigated these approaches by implementing two different strategies: L1 logistic regression (L1LR) that models a single global solution and classification trees that model multiple local solutions corresponding to discriminant subregions of the feature space. For each strategy, we built a standard model, and a sparser version of it. As an alternative to pruning, we explore a promising approach that first constrains the tree model with an L1LR-based feature selection, an approach we called L1LR-Tree. The objective is to decrease its vuln...
... Views on the choice of simulation tools, compared in the context of the same model: the bacte... more ... Views on the choice of simulation tools, compared in the context of the same model: the bacteriophage Lambda. Anastasia Yartseva 1 , Denis Mestivier 2 , Pierre-Yves Boëlle 3 , Adrien Richard 1 , Guillaume Hutzler 1. (2005). ...
Metabolic profiling, the study of changes in the concentration of the metabolites in the organism... more Metabolic profiling, the study of changes in the concentration of the metabolites in the organism induced by biological differences within subpopulations, has to deal with a very large amount of complex data. It therefore requires the use of powerful data processing and machine learning methods. To overcome over-fitting, a common concern in metabolic profiling where the number of features is often much larger than the number of observations, many predictive analyses combined dimension reduction techniques with multivariate predictive linear modeling. Moreover, they built a global model that identifies biomarkers predictive of the output of interest giving their overall trend variations. However, this fails to capture local biological phenomena underlying subgroups of subjects. More recently, local exploration methods based on decision trees approaches have been applied in metabolomics but they only explore random parts of the feature space. In this study, we used a supervised rule-m...
... Ateliers de démonstration et de comparaison d'outils de modélisation autour de modèles b... more ... Ateliers de démonstration et de comparaison d'outils de modélisation autour de modèles biologiques communs. Anastasia Yartseva 1 , Denis Mestivier 2 , Pierre-Yves Boëlle 3 , Adrien Richard 1 , Guillaume Hutzler 1. (2005). ...
Uploads
Papers by Anastasia Yartseva