Toward unbiased objective functions in constraint-based modeling
Other text in scientific journal, 2026
The objective function is important to constraint-based modeling, but setting it has traditionally been subjective, limiting predictive accuracy. We expose the biases of commonly used objective functions and propose a paradigm shift toward unbiased objective functions powered by artificial intelligence, offering a promising avenue for metabolic engineering and biomedical applications.
artificial intelligence
constraint-based model
objective function
Author
Weihang Dong
Shenyang University of Chemical Technology
Shenzhen Institute of Advanced Technology
Jens B Nielsen
BioInnovation Institute
Chalmers, Life Sciences, Systems and Synthetic Biology
Haiyan Li
Shenyang University of Chemical Technology
Feiran Li
Tsinghua University
Yu Chen
Shenzhen Institute of Advanced Technology
Trends in Biotechnology
0167-7799 (ISSN) 18793096 (eISSN)
Vol. In PressSubject Categories (SSIF 2025)
Clinical Medicine
Computer and Information Sciences
DOI
10.1016/j.tibtech.2026.01.004