Fluxomics workflow: case studies

By Carles Foguet, Effrosyni Karakitsou, Pedro de Atauri and Marta Cascante (last updated: 2018-07-28)

The following table contains case studies from literature and databases ¹³C tracer data based on mass spectrometry, which are applied for estimation of internal fluxes according to the principles of ¹³C Metabolic Flux Analysis (¹³C-MFA). Each case corresponds to our free interpretation and adaptation of the data provided by the original authors. The different case-studies are organized around three models of central metabolism:

• 1. Mammals (human and hamster).
• 2. Yeast (Sacharomyces cerevisiae).
• 3. Bacteria (Escherichia coli).

These models are provided as templates to be used in new analyses. They follow the nomenclature conventions in the BIGG Models data base for genome-scale models: Recon3D for human; iCHOv1 for hamster; iMM904 for Sacharomyces cerevisiae; and iML1515 for Escherichia coli. This permits the integration of ¹³C-MFA studies with genome-scale models, by adding the results of ¹³C-MFA as flux constraints to genome-scale metabolic models.

Reference		Organism/Tissue/Cell line	13C data	Labelled substrates and products	Suggested escher map	Required input files
Metabolights MTBLS412	-	Human HUVEC cells (hypoxia)	Raw data	Glc to Rib, Glycogen, Lac	map	files
PMID29922569	doi	Human MCF-7 breast carcinoma cells (Control)	Corrected data	Gln, Glc to 3PG, Asp, Pyr, Ala	map	files
PMID27110360	doi	Human lung adenocarcinoma A549 cells (21%O2)	Corrected data	Glc, Gln to Mal, N-acetyl-Asp, Cit, AMP, Glu	map	files
PMID27379180	doi	Human pancreatic ductal adenocarcinoma PANC1 cell line (TNF-alpha)	Corrected data	Glc to Pyr, Cit	map	files
PMID24773761	doi	Cricetulus griseus CHO-K1 cell line (Compartimentation)	Corrected data	Glc to Asp, Glu, Gly, Ser	map	files
PMID27761435	doi	Saccharomyces cerevisiae WRY2 (Base strain - glucose)	Corrected data	Glc to Gly, Ala, Val, Thr, Leu, Ile, Asn, Glu, Gln, Arg, Phe, Tyr	map	files
PMID16269086	doi	Saccharomyces cerevisiae (Purely oxidative growth)	Corrected data	Glc to Ala, Gly, Ser, Thr, Asp, Arg, Phe, Tyr, Glu	map	files
PMID29142823	doi	Escherichia coli K12 MG1655 ([0:8:2])	Corrected data	Glc to Ala, Ser, Thr, Tyr	map	files
PMID24957033	doi	Escherichia coli K12 MG1655 (FAAs)	Corrected data	Glc to Ala, Asp, Glu, Phe, Gly, Val, Tyr, Thr	map	files

The files required to run each example in the workflow for fluxomics are provided in the last column. There are several possible workflows depending on the selected combination of programs. Among them, two ways of applying the workflow for fluxomics:

• Starting with raw data. This corresponds to the example used in the fluxomics workflow tutorial and includes a complete analysis using Ramid, [Midcor](https://github.com/phnmnl/phenomenal-h2020/wiki/Midcor), [Iso2flux](https://github.com/cfoguet/iso2flux) and [escher-fluxomics](https://github.com/phnmnl/container-escher-fluxomics).
• Starting with corrected data obtained from literature. The applications of Ramid and Midcor are not required, and the workflow must be edited to be like this:

For visualizations of the maps of central metabolism, escher-fluxomics take advantage of the web-based tool Escher and maps available in the community-maps repository for Escher.

Iso2flux is based on a heuristic optimization and, accordingly, complex systems as those presented here might require to run several times the workflow with Iso2flux until a succesful fitting is obtained

Acknowledgements

Carles Foguet, Effrosyni Karakitsou, Pedro de Atauri and Marta Cascante (Universitat de Barcelona); Pablo Moreno and Namrata Kale (EMBL-EBI). The data covering these case studies were first collected by Arathi Raghunath and Usha Mahadevan (Molecular Connections) and then interpreted and adapted by Pedro de Atauri.