Search

The selection of informative variables in partial least squares (PLS) is important in process analytical technology (PAT) applications in the pharmaceutical industry, for example, the calibration of spectrometers. In the past, numerous approaches have been proposed to select the variables in partial least squares. In this work, a new variable selection method for PLS with the weighted variable contribution (PLS-WVC) to the first singular value of the covariance matrix for each PLS component is proposed. Several variants of PLS-WVC with different weighting factors are proposed. One variant of PLS-WVC is equivalent to the PLS with variable importance in projection (PLS-VIP). However, the variants with the correlation between X γ w γ and Y γ q γ as the weighting factor are preferred based on the results of the simulation cases studies. The proposed PLS-WVCs are integrated with interval PLS (iPLS) further to select the informative wavelength intervals for spectroscopic modelling. The utility of the proposed WVC based variable selection methods in PLS is demonstrated with the real spectral data sets.

N -propanol is an important precursor in industrial antibiotic erythromycin fermentation by Saccharopolyspora erythraea . However, the online control of n -propanol concentration is a challenging problem, and hardly any effective methods have been applied in the fermentation process. In this study, an innovative method developed for on-line detection and feedback control of n -propanol concentration in erythromycin fermentation process with electronic nose. The models for n -propanol detection based on the response of the electronic nose were established, and the on-line monitoring feedback system for n -propanol concentration controlling strategy was applied in erythromycin fermentation. The effects of different residual n -propanol concentration in fermentation broth on physiological metabolism of S. erythraea were investigated in a 50-L bioreactor. The results revealed that maintaining an n -propanol concentration of 1000 ± 38 ppm, detected by electronic nose, resulted in the highest erythromycin yield (8500 U/mL). In conclusion, real-time detection and feedback control of n -propanol concentration using electronic nose is an effective method for erythromycin fermentation processes and a new precursor adding optimization process was developed.

The isotopically non-stationary 13C labelling experiments, as an emerging experimental technique, can estimate the intracellular fluxes of the cell culture under an isotopic transient period. However, to the best of our knowledge, the issue of the structural identifiability analysis of non-stationary isotope experiments is not well addressed in the literature. In this work, the local structural identifiability analysis for non-stationary cumomer balance equations is conducted based on the Taylor series approach. The numerical rank of the Jacobian matrices of the finite extended time derivatives of the measured fractions with respect to the free parameters is taken as the criterion. It turns out that only one single time point is necessary to achieve the structural identifiability analysis of the cascaded linear dynamic system of non-stationary isotope experiments. The equivalence between the local structural identifiability of the cascaded linear dynamic systems and the local optimum condition of the nonlinear least squares problem is elucidated in the work. Optimal measurements sets can then be determined for the metabolic network. Two simulated metabolic networks are adopted to demonstrate the utility of the proposed method.

Background Pichia pastoris is one of the most important cell factories for production of industrial enzymes and heterogenous proteins. The genome-scale metabolic model of high quality is crucial for comprehensive understanding of the P. pastoris metabolism. Methods In this paper, we upgraded P. pastoris genome-scale metabolic model based on the combination of latest genome annotations and literatures. Then the performance of the new model was evaluated using the Cobra Toolbox v2.0. Results Compared with the recently published model iMT1026, the reaction number in the new model iRY1243 was increased from 2035 to 2407 and the metabolite number was increased from 1018 to 1094. Accordingly, the unique ORF number was increased from 1026 to 1243. To improve the metabolic functions of P. pastoris genome-scale metabolic model, the biosynthesis pathways of vitamins and cofactors were carefully added. iRY1243 showed good performances when predicting the growth capability on most of the reported carbon and nitrogen sources, the metabolic flux distribution with glucose as a sole carbon source, the essential and partially essential genes, and the effects of gene deletion or overexpression on cell growth and S-adenosyl-l-methionine production. Conclusion iRY1243 is an upgraded P. pastoris genome-scale metabolic model with significant improvements in the metabolic coverage and prediction ability, and thus it will be a potential platform for further systematic investigation of P. pastoris metabolism. Electronic supplementary material The online version of this article (doi:10.1186/s40643-017-0152-x) contains supplementary material, which is available to authorized users.

Pichia pastoris is a popular recombinant protein expression system for its accessibility of efficient gene manipulation and high protein production. Sufficient supply of precursors, energy, and redox cofactors is crucial for high recombinant protein production. In our present work, we found that the addition of glutamate improved the recombinant β-galactosidase (β-gal) production by P. pastoris G1HL. To elucidate the impacts of glutamate on the central metabolism in detail, a combined 13C-assisted metabolomics and 13C metabolic flux analysis was conducted based on LC–MS/MS and GC–MS data. The pool sizes of intracellular amino acids were obviously higher on glucose/glutamate (Glc/Glu). The fluxes in EMP entry reaction and in downstream TCA cycle were 50 and 67% higher on Glc/Glu than on Glc, respectively. While the fluxes in upstream TCA cycle kept almost unaltered, the fluxes in PPP oxidative branch decreased. The addition of glutamate leads to a remarkable change on the central metabolism of high β-galactosidase-producing P. pastoris G1HL. To meet the increased demands of redox cofactors and energy for higher β-galactosidase production on Glc/Glu, P. pastoris G1HL redistributes the fluxes in central metabolism through the inhibitions and/or activation of the enzymes in key nodes together with the energy and redox status.