Integrative description among <i>in silico</i> analysis and high-throughput data.

<p>(A) Log-ratio and <i>p</i>-values plot. In blue we show those metabolites that obeyed two of three criteria: a log-ratio higher than 2, a <i>p</i>-value lower than 0.01,and its requirement for consideration as a metabolite with an important role in supporting bacterial nitrogen fixation. (B) Historical improvements in OFs. Since the first publication dealing with the systems biology description of the bacterial nitrogen fixation for <i>R. etli </i><a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1002720#pcbi.1002720-ResendisAntonio3" target="_blank">[19]</a>, some improvements have been carried out for determining the OF, by identifying proper metabolites required for supporting this biological process (see <a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1002720#pcbi-1002720-g002" target="_blank">Figure 2</a>). At present, the suggested objective function is the result of a careful search of the published literature and an integrative description in the metabolome data. Under these experimental and physiological bases, our modeling located an OF slightly closer to the complex metabolism supporting bacterial nitrogen fixation, and consequently constraint-based modeling raised its accuracy and predictive scope. D) Comparative analysis between FBA and transcriptome and proteome data. A detailed comparison between computational predictions and high-throughput data was carried out for all 22 metabolic pathways defined in panel C. One can distinguish three possible situations: 1) genes (enzymes) that were predicted <i>in silico</i> but not monitored experimentally (CBM not HT; blue); 2) genes (enzymes) that were consistently observed in both schemes (CBM and HT; green); and 3) genes (enzymes) that were experimentally monitored but not predicted <i>in silico</i> (HT not CBM; red). As explained in the <a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1002720#s4" target="_blank">Materials and Methods</a> section, <i>η</i> is related to the fraction of genes (enzymes) that were consistently observed in both schemes and constitutes the backbone for our modeling assessment. Upper and lower bars at the right side of panel D indicate the fraction of proteins and genes that were predicted <i>in silico</i> and detected in proteome or transcriptome data.