Recent advances in cell genome editing techniques enable the generation of high-throughput gene knockout data in the malaria parasites in vivo. Integrative analysis of this data can lead to the identification of biological mechanisms that explain the observed phenotypes and that provide testable hypotheses for further discoveries. Metabolic modelling can cope with the tangled metabolism of the malaria parasites, and hence is a compelling approach for understanding the parasites physiology. In this study, we present a combined experimental and computational approach that suggests cellular mechanisms for targeting the malaria parasites. We predict in silico and test in vivo lethal knockouts and synthetic lethal pairs in the blood and liver stages of the malaria infection. We perform computational analyses on a newly developed genome-scale model of the malaria parasite Plasmodium berghei (iPbe), and we use high-throughput gene knockout data generated in the PlasmoGEM project. The comparison between data and gene essentiality predictions allow the understanding of the parasite’s physiology in the blood and liver stages. We identify the thermodynamic bottlenecks, genetic interactions, and the accessibility to nutrients behind the phenotypes. When we simulate in iPbe the hypothesised physiology, we achieve 80% consistency between the prediction of essential genes and the experimental data. This result indicates that our model iPbe is a valuable framework for the generation of testable hypothesis on antimalarial targets. Overall, the knowledge generated in this framework will serve to tackle more efficiently the malaria parasites’ metabolism during infection.