Assessment of transcriptomic constraint-based methods for central carbon flux inference

Area of Science:

• Systems Biology
• Metabolic Engineering
• Computational Biology

Background:

• Intracellular metabolic flux determination using 13C metabolic flux analysis (13C-MFA) is resource-intensive.
• Previous studies suggested transcriptomic data combined with constraint-based metabolic modeling (CBM) can predict fluxes accurately.
• Prior validation was limited to specific organisms and conditions, necessitating broader assessment.

Purpose of the Study:

• To extensively validate the utility of transcriptomic data coupled with CBM for intracellular flux prediction across diverse microbial metabolisms.
• To compare the performance of different computational flux-balance analysis (FBA) methods.
• To assess the impact of known versus unknown substrate uptake rates on prediction accuracy.

Main Methods:

• Compiled a dataset of transcriptomic and 13C-MFA flux data for 21 experimental conditions across various unicellular organisms and substrates.
• Assessed three computational FBA methods: E-Flux2, SPOT, and pFBA.
• Evaluated prediction accuracy under conditions with known and unknown carbon source and metabolite uptake rates.

Main Results:

• When uptake rates were known, transcriptomic data offered no significant advantage over non-transcriptomic CBM (pFBA).
• When uptake rates were unknown, transcriptomic data integration significantly improved flux prediction accuracy compared to CBM alone.
• Transcriptomic-informed methods (E-Flux2, SPOT) showed substantially higher correlation coefficients than pFBA in unknown uptake rate scenarios.

Conclusions:

• Transcriptomic data coupled with CBM is a powerful approach for estimating intracellular metabolic fluxes in microorganisms.
• This integrated method is particularly valuable when substrate uptake rates are difficult to determine, such as in complex media or in vivo.
• The findings support the broader applicability of transcriptomic-informed metabolic modeling beyond simple laboratory conditions.