, Equations 199 (3) to (6) set constraints corresponding to gene expression data. y i + , y i-and y i are boolean variables 200 representing the adequacy between the predicted flux ? i through reaction i and its expression level

, & (4) enforce that, if y i + =1 or y i-=1 (i.e., the reaction is active), the value 203 of the flux through reaction i is larger than a threshold ?, whereas if y i + =0 or y i-=0 (i.e., the reaction 204 is inactive), the flux through reaction i must be 0. For NEr, y i represents whether the reaction is 205 inactive: equations (5) & (6) enforce that, For HEr, y i + (y i-) represents whether the reaction is active in the forward (or backward, respectively) 202 direction: equations (3)

S. , respectively) are Boolean variable representing whether the reaction i is active or not 211 in the forward (or backward, respectively) direction. When x i + = 1, the flux through reaction i must 212 be larger (or smaller, respectively) than ? (equation

, If x i + =0 and x i-=0, the flux through 215 reaction i is forced to be 0, forcing the reaction to be inactive. The constraints set in Eq. 9 ensure 216 that at least one reaction is able to produce the metabolite j is active. 217 The optimization problem consists in finding the flux distribution ?, which maximizes the 218 number of HEr which are active (? HEr ? ?) and the number of NEr which are inactive (? NEr = 0). the reactions were classified as "required" (R) if they were found to be active in all solutions 686 and conversely "Inactive" (I) if they display a zero-flux in all the optimal solutions. All other 687 reactions, found to be either active or inactive in the optimal solutions, were classified as 688 "Potentially active" (PA). This analysis was made independently for the d3 and d30 differentiation 689 stages. (B) Reactions identified as I at d3 but R or PA at d30 were considered as activated, For reversible reactions, when x i-= 1, the flux through reaction i must be negative and lower than 214 ? (equation (8)), forcing the reaction to be backward active

. Chazalviel, Principal component analysis of 3-day and 30-day subnetworks. PCA analysis was the MetExplore webserver, 2018.

, Results from liver-specific metabolic functions simulated for each of the cell-specific models 727 identified at each stage: for each type of function

, Assessment of the liver-specificity of generated 740 subnetworks: comparison with UNIPROT data (Figure S7); Assessment of the liver-specificity of 741 generated subnetworks: comparison with Human Proteome Atlas data (Figure S8); made from conjoint or individual analysis of cell-specific models for the keratin sulfate 747 degradation pathway (Figure S12); the Ubiquinone, N-glycan synthesis 753 and keratin sulfate degradation pathways (Supplementary Notes) (PDF) 754 Metabolic functions used for simulations (Table S1, XLSX) 755 NMR identified metabolites for 3-day and 30, Principle of the iMat algorithm illustrated on a Toy model

, Computed subnetworks for 3-day and 30-day HepaRG cells from our method and the 757 FASTCORE algorithm (Table S4, XLSX)

, This material is available free of charge via the

?. Srsmc, ?INSB Institut des sciences biologiques, 764 UMR8204 Centre d'infection et d'immunité de

F. J. , N. P. , A. C. , F. Fmo, H. et al.,

C. F. Fma, ;. Fj, D. Z. Bf, and N. C. , FV participated in the computational analysis; NP wrote the manuscript, NP contributed in data interpretation, discussion on results and writing and revising the manuscript

, This work was supported by the French Ministry of Research and National Research Agency as 772 part of the French MetaboHUB, the national metabolomics and fluxomics infrastructure

, This work was supported in part by the French Ministry of Research and the National Research

, Agency, as part of the French MetaboHUB infrastructure (the national metabolomics and 778 fluxomics infrastructure, Grant ANR-INBS-0010) and the NISTEC project

I. Thiele and B. Ø. Palsson, Reconstruction Annotation Jamborees: A Community Approach to 784 Systems Biology, Mol. Syst. Biol, vol.6, p.361, 2010.

N. C. Duarte, S. A. Becker, N. Jamshidi, I. Thiele, M. L. Mo et al., Global Reconstruction of the Human Metabolic Network Based on Genomic 787 and Bibliomic Data, Proc. Natl. Acad. Sci. U. S. A, vol.786, issue.6, pp.1777-1782, 2007.

H. Ma, A. Sorokin, A. Mazein, A. Selkov, E. Selkov et al., Goryanin, I. The 789 Edinburgh Human Metabolic Network Reconstruction and Its Functional Analysis. Mol Syst 790 Biol, vol.3, 2007.

I. Thiele, N. Swainston, R. M. Fleming, A. Hoppe, S. Sahoo et al., Metabolic Modelling of Hepatocytes Reveals Serine Deficiency in Patients with Non793 Alcoholic Fatty Liver Disease, vol.5, 2013.

, Global Reconstruction of Human Metabolism, Nat Biotechnol, vol.31, issue.5, pp.419-425, 2013.

K. Smallbone, N. Swainston, K. Smallbone, H. Hefzi, P. D. Dobson et al., Striking a Balance with Recon 2.1, vol.798, 2013.

D. C. Ang, K. S. Gardiner, N. J. Gutierrez, and J. M. , Recon 2.2: From Reconstruction 800 to Model of Human Metabolism, Metabolomics, vol.12, issue.7, p.109, 2016.

J. A. Papin, E. Brunk, S. Sahoo, D. C. Zielinski, A. Altunkaya et al., Recon3D Enables a Three-Dimensional 806 View of Gene Variation in Human Metabolism, Reconciled Rat and Human Metabolic Networks for Comparative 803 Toxicogenomics and Biomarker Predictions, vol.8, pp.272-281, 2017.

T. Shlomi, M. N. Cabili, M. J. Herrgard, B. Ø. Palsson, E. Ruppin et al., , vol.808

, Metabolism. Nat Biotech, vol.26, issue.9, pp.1003-1010, 2008.

L. Jerby, T. Shlomi, and E. Ruppin, Computational Reconstruction of Tissue-Specific, vol.810, issue.11

A. Bordbar, N. Jamshidi, and B. O. Palsson, Metabolic Models: Application to Human Liver Metabolism, vol.6, 2010.

R. Agren, S. Bordel, A. Mardinoglu, N. Pornputtapong, I. Nookaew et al., Knowledge-Base of Erythrocyte Metabolism That Can Be Used to Simulate Its 814 Physiological and Patho-Physiological States, PLoS Comput Biol, vol.5, issue.13, p.1002518, 2011.

R. Agren, A. Mardinoglu, A. Asplund, C. Kampf, M. Uhlen et al., , p.819

, Anticancer Drugs for Hepatocellular Carcinoma through Personalized Genome-Scale

, Metabolic Modeling. Mol. Syst. Biol, vol.10, pp.1-13, 2014.

C. Frezza, L. Zheng, O. Folger, K. N. Rajagopalan, E. D. Mackenzie et al.,

, Lethal with the Tumour Suppressor Fumarate Hydratase, Nature, vol.477, issue.7363, pp.225-824, 2011.

I. Goldstein, K. Yizhak, S. Madar, N. Goldfinger, E. Ruppin et al., P53 Promotes 826 the Expression of Gluconeogenesis-Related Genes and Enhances Hepatic Glucose 827 Production, Cancer Metab, vol.825, issue.16, p.9

K. Yizhak, S. E. Le-dévédec, V. M. Rogkoti, F. Baenke, V. C. De-boer et al., , p.829

A. Schulze, B. Van-de-water, and E. Ruppin, A Computational Study of the Warburg Effect 830 Identifies Metabolic Targets Inhibiting Cancer Migration, Mol. Syst. Biol, vol.10, p.744, 2014.

J. D. Orth, I. Thiele, and B. Ø. Palsson, What Is Flux Balance Analysis?, Nat Biotechnol, vol.832, issue.3, pp.245-248, 2010.

H. Zur, E. Ruppin, T. Shlomi, and . Imat, An Integrative Metabolic Analysis Tool. 834 Bioinformatics, vol.26, pp.3140-3142, 2010.

C. Aninat, A. Piton, D. Glaise, T. Le-charpentier, S. Langouët et al.,

C. Guillouzo and A. Guillouzo, Expression of Cytochrome P450 Enzymes and Nuclear 837 Receptors in Human Hepatoma HepaRG Cells, Drug Metab. Dispos, vol.34, issue.1, pp.75-83, 2006.

R. M. Rodrigues, A. Heymans, V. De-boe, A. Sachinidis, U. Chaudhari et al., Toxicogenomics-Based Prediction of, vol.838

, Acetaminophen-Induced Liver Injury Using Human Hepatic Cell Systems, Toxicol. Lett, vol.841, issue.1, pp.50-59, 2016.

L. Peyta, K. Jarnouen, M. Pinault, C. Guimaraes, and S. Chevalier, J.-P, vol.842, issue.22

, Decreases Respiratory Chain Capacities and Increases ATP Synthesis Yield in the Human

, HepaRG Cells. Biochim. Biophys. Acta-Bioenerg, vol.1857, issue.4, pp.443-453, 2016.

X. Ma, Y. Duan, B. Tschudy-seney, G. Roll, I. S. Behbahan et al., Highly Efficient Differentiation of Functional, vol.848

, Hepatocytes From Human Induced Pluripotent Stem Cells, Stem Cells Transl. Med, vol.2, issue.6, pp.409-419, 2013.

J. Zhang, E. Nuebel, G. Q. Daley, C. M. Koehler, M. A. Teitell et al., , p.851

, Pluripotent Stem Cells during Reprogramming and Self-Renewal. Cell Stem Cell, vol.11, issue.5, pp.589-595, 2012.

X. Xu, S. Duan, F. Yi, A. Ocampo, G. Liu et al., Mitochondrial 854 Regulation in Pluripotent Stem Cells, Cell Metab, vol.18, issue.3, pp.325-332, 2013.

T. Teslaa and M. A. Teitell, Pluripotent Stem Cell Energy Metabolism: An Update, EMBO J, vol.856, issue.2, pp.138-153, 2015.

S. Bucher, P. Jalili, D. Le-guillou, K. Begriche, K. Rondel et al., Bisphenol a Induces Steatosis in HepaRG Cells 859 Using a Model of Perinatal Exposure, Environ. Toxicol, vol.2017, issue.3, pp.1024-1036

V. Cerec, D. Glaise, D. Garnier, S. Morosan, B. Turlin et al., Transdifferentiation of Hepatocyte-like 862 Cells from the Human Hepatoma HepaRG Cell Line through Bipotent Progenitor, vol.861

, Hepatology, vol.45, pp.957-967, 2007.

, Whole Genome Gene Expression Profiles of HepaRG Cells and HepG2 Cells to Primary

, Human Hepatocytes and Human Liver Tissues, Drug Metab. Dispos, issue.6, pp.988-994, 2010.

R. Parent and L. Beretta, Translational Control Plays a Prominent Role in the Hepatocytic 868 Differentiation of HepaRG Liver Progenitor Cells, Genome Biol, vol.867, issue.30, p.19, 2008.

D. S. Wishart, T. Jewison, A. C. Guo, M. Wilson, C. Knox et al., , vol.870

R. Mandal, F. Aziat, and E. Dong, HMDB 3.0-The Human Metabolome Database in 871 2013, Nucleic Acids Res, issue.D1, pp.801-807, 2012.

S. Rossell, M. A. Huynen, and R. A. Notebaart, Inferring Metabolic States in Uncharacterized 873 Environments Using Gene-Expression Measurements, PLoS Comput Biol, vol.2013, issue.3, p.1002988

, UniProt: The Universal Protein Knowledgebase, Nucleic Acids Res, vol.875, issue.33, pp.158-876, 2017.

M. Uhlen, L. Fagerberg, B. M. Hallstrom, C. Lindskog, P. Oksvold et al., , vol.878

, Proteome. Science, issue.6220, pp.1260419-1260419, 2015.

G. Dennis, B. T. Sherman, D. A. Hosack, J. Yang, W. Gao et al., Database for Annotation, vol.881, p.882

, Visualization, and Integrated Discovery, vol.4, p.60, 2003.

A. Bordbar, A. M. Feist, R. Usaite-black, J. Woodcock, B. O. Palsson et al., A 884 Multi-Tissue Type Genome-Scale Metabolic Network for Analysis of Whole-Body Systems

C. Gille, C. Bolling, A. Hoppe, S. Bulik, S. Hoffmann et al., HepatoNet1: A Comprehensive Metabolic 888 Reconstruction of the Human Hepatocyte for the Analysis of Liver Physiology, Mol Syst, vol.887, issue.411, p.411, 2010.

L. Cottret, D. Wildridge, F. Vinson, M. P. Barrett, H. Charles et al., 891 MetExplore: A Web Server to Link Metabolomic Experiments and Genome-Scale

, Metabolic Networks, Nucleic Acids Res, vol.38, pp.132-139, 2010.

N. Vlassis, M. P. Pacheco, and T. Sauter, Fast Reconstruction of Compact Context-Specific 894

, Metabolic Network Models, vol.10, pp.1-15, 2013.

L. Tong, Structure and Function of Biotin-Dependent Carboxylases, Cell. Mol. Life Sci, vol.896, issue.5, pp.863-891, 2013.
DOI : 10.1007/s00018-012-1096-0
URL : http://europepmc.org/articles/pmc3508090?pdf=render

, 897 (41) Badawy, A. A.-B. Tryptophan Metabolism, Disposition and Utilization in Pregnancy

, Biosci. Rep, vol.35, issue.5, pp.261-00261, 2015.

T. Suda, J. C. Robinson, and T. A. Fjellstedt, Developmental Changes in the Enzymatic 900 Capacity for Reduction and Oxidation of Alpha-Ketoadipate in Rat Liver, Heart, Kidney, 901 and Brain, Pediatr. Res, vol.12, issue.4, pp.297-300, 1978.

M. Chazalviel, C. Frainay, N. Poupin, F. Vinson, B. Merlet et al., MetExploreViz: Web Component for Interactive Metabolic Network 904 Visualization, Bioinformatics, vol.34, issue.43, pp.312-313, 2018.
DOI : 10.1093/bioinformatics/btx588
URL : https://academic.oup.com/bioinformatics/article-pdf/34/2/312/25114213/btx588.pdf

R. J. Wanders, H. R. Waterham, S. Ferdinandusse, and . Metabolic, , p.905

, Endoplasmic Reticulum. Front. Cell Dev. Biol, vol.3, p.83, 2016.

S. Kersten, Integrated Physiology and Systems Biology of PPAR?, Mol. Metab, vol.2014, issue.4, pp.909-354
DOI : 10.1016/j.molmet.2014.02.002
URL : https://doi.org/10.1016/j.molmet.2014.02.002

C. Y. Fan, J. Pan, R. Chu, D. Lee, K. D. Kluckman et al.,

M. S. Rao and N. Maeda, Hepatocellular and Hepatic Peroxisomal Alterations in Mice 912 with a Disrupted Peroxisomal Fatty Acyl-Coenzyme A Oxidase Gene, J. Biol. Chem, vol.913, issue.40, pp.24698-24710, 1996.

J. C. Newman, E. Verdin, and . ?-hydroxybutyrate, A Signaling Metabolite, Annu. Rev. Nutr, vol.915, issue.1, pp.51-76, 2017.

A. Zelezniak, T. H. Pers, S. Soares, M. E. Patti, K. R. Patil et al., Metabolic Network Topology 917 Reveals Transcriptional Regulatory Signatures of Type 2 Diabetes, Enrichment Analysis of Metabolomic Data. 920 Bioinformatics, vol.918, pp.730-731, 2010.

M. Persicke, C. Rückert, J. Plassmeier, L. J. Stutz, N. Kessler et al., Metabolite Set Enrichment Analysis in the MeltDB Metabolomics 923 Software Platform: Metabolic Profiling of Corynebacterium Glutamicum as an Example

, Metabolomics, vol.2012, issue.2, pp.310-322

C. Frainay and F. Jourdan, Computational Methods to Identify Metabolic Sub-Networks Based 926 on Metabolomic Profiles, Brief. Bioinform, vol.2017, issue.1, pp.43-56