TY - JOUR
T1 - Evaluation of an S-system root-finding method for estimating parameters in a metabolic reaction model
AU - Iwata, Michio
AU - Miyawaki-Kuwakado, Atsuko
AU - Yoshida, Erika
AU - Komori, Soichiro
AU - Shiraishi, Fumihide
N1 - Funding Information:
The authors would like to thank Professor Eberhard O. Voit at Georgia Institute of Technology for his theoretical suggestions. This work was supported by JSPS Science Grant-in-Aid for JSPS Research Fellow Grant Number 17J04007.
Funding Information:
The authors would like to thank Professor Eberhard O. Voit at Georgia Institute of Technology for his theoretical suggestions. This work was supported by JSPS Science Grant-in-Aid for JSPS Research Fellow Grant Number 17J04007.
PY - 2018/7
Y1 - 2018/7
N2 - In a mathematical model, estimation of parameters from time-series data of metabolic concentrations in cells is a challenging task. However, it seems that a promising approach for such estimation has not yet been established. Biochemical Systems Theory (BST) is a powerful methodology to construct a power-law type model for a given metabolic reaction system and to then characterize it efficiently. In this paper, we discuss the use of an S-system root-finding method (S-system method) to estimate parameters from time-series data of metabolite concentrations. We demonstrate that the S-system method is superior to the Newton–Raphson method in terms of the convergence region and iteration number. We also investigate the usefulness of a translocation technique and a complex-step differentiation method toward the practical application of the S-system method. The results indicate that the S-system method is useful to construct mathematical models for a variety of metabolic reaction networks.
AB - In a mathematical model, estimation of parameters from time-series data of metabolic concentrations in cells is a challenging task. However, it seems that a promising approach for such estimation has not yet been established. Biochemical Systems Theory (BST) is a powerful methodology to construct a power-law type model for a given metabolic reaction system and to then characterize it efficiently. In this paper, we discuss the use of an S-system root-finding method (S-system method) to estimate parameters from time-series data of metabolite concentrations. We demonstrate that the S-system method is superior to the Newton–Raphson method in terms of the convergence region and iteration number. We also investigate the usefulness of a translocation technique and a complex-step differentiation method toward the practical application of the S-system method. The results indicate that the S-system method is useful to construct mathematical models for a variety of metabolic reaction networks.
UR - http://www.scopus.com/inward/record.url?scp=85044366891&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85044366891&partnerID=8YFLogxK
U2 - 10.1016/j.mbs.2018.01.010
DO - 10.1016/j.mbs.2018.01.010
M3 - Article
C2 - 29410225
AN - SCOPUS:85044366891
SN - 0025-5564
VL - 301
SP - 21
EP - 31
JO - Mathematical Biosciences
JF - Mathematical Biosciences
ER -