The overall objective of this work was to reconstruct and perform in silico phenotype simulations for integrated models of metabolism and regulation. The number of genomes available in the public domain increased exponentially in the last decade. The overwhelming amount of data led to the introduction of automated pipelines for genome annotation, also facilitating the propagation of annotation inconsistencies from public repositories. In this work, we explore the use of GEMs as tools for annotation curation. A protocol for annotation curation with metabolic network reconstructions was designed and applied to the genus Brucella. The high-throughput reconstruction and analysis of genome-scale transcriptional regulatory networks is a current challenge in SB research. In this work, the model organism Bacillus subtilis was chosen as a case study and a new manually curated network for its transcriptional regulation was introduced. We proposed a new methodology for the inference of regulatory interactions from gene expression data. The newly proposed methodology dubbed “atomic regulon inference” was shown to capture many sets of genes corresponding to regulatory units in the manually curated network.
 
Following this line of work, based on the proposed regulatory transcriptional regulatory network for B. subtilis, we introduced an integrated genome-scale model for the metabolism and transcriptional regulation in B. subtilis. Model validation was performed with in silico growth phenotype simulations for mutant strains described in the literature. The integrated model was able to predict transcription factor knockouts for growth in multiple environmental conditions, expanding the predictive capabilities of the metabolic model by itself.