*Trans-Flux: An automated pipleline for integrating transcriptomic data with Flux Balance Analysis (FBA) simulations*
Recent advances in genomic sequencing technologies have enabled reconstruction of genome scale metabolic network models. These genome scale metabolic models have been employed for metabolic phenotype prediction, metabolic engineering, drug target identification etc. Metabolic Flux balance analysis (FBA) is one of the most used constraint based technique applied to metabolic models, for the prediction of metabolic phenotypes. However, the predictions of FBA, which primarily rely on basic mass balance constraints, often results in 'outlier' and 'biologically irrelevant' fluxes, which can be attributed to the existance of multiple LP solutions to a given optimiztion problem. Some recent approaches have utilized gene expression data for defining reaction constraints in order to obtain results which would be coherent with biological observations (Becker and Palsson, 2008; van Berlo et al., 2011; Colijn et al., 2009; Song et al., 2014; Zur et al., 2010). These methods work either by constraining the bounds of individual reactions, or by completely inactivating them, based on a defined expression-threshold of the corresponding genes/ enzymes (Becker and Palsson, 2008; Colijn et al., 2009). However, it may be noted that results of the simulations are greatly dependent on the basis of constraints, and therefore needs to be carefully chosen (Raman and Chandra, 2009). Utilizing gene and protein expression to predict metabolic flux is a challenging task not only due to the complex mapping between genes/proteins and reactions but also due to the computationally expensive algorithms.
Here, we present TransFlux, a stand-alone tool which allows integration of gene/protein expression data with a genome scale metabolic model. This tool works in any Linux/Unix environment. TransFlux employs a novel yet simple algorithm for constraining fluxes and providing optimal fast and accurate solutions. This tool has been validated using gene expression and metabolite data of E. coli, and could outperform currently available techiques/ algorithms for transcriptomic data integration. Furthermore, it may also be noted that none of the currently available algorithms come with user friendly implementations. Thus, we expect that TransFlux will bridge this gap and enable the scientific community to better investigate the cellular metabolic potential and its dependencies on the transcriptome/gene-regulation. We provide below a brief description of the Transflux tool with details of input/Outputs.