مقالات پذیرفته شده در نهمین کنگره بین المللی زیست پزشکی
Integrating Transcriptomic Data into Flux Balance Analysis: Algorithms, Challenges, and Biomedical Applications
Integrating Transcriptomic Data into Flux Balance Analysis: Algorithms, Challenges, and Biomedical Applications
Sayyed Amir Sarabi,1,*
1. Independent Researcher, currently studying at Iran University of Science and Technology
Introduction: Integrating transcriptomic data into genome-scale metabolic models enhances the predictive accuracy of cellular metabolism and supports biomedical applications such as disease modeling, metabolic engineering, and drug target identification. Various computational algorithms have been developed to incorporate omics data into flux balance analysis (FBA), each with distinct methodologies, strengths, and limitations.
Methods: GIMME integrates transcriptomic data with FBA to produce condition-specific metabolic networks, identifying inactive reactions while preserving essential metabolic functions. TRFBA incorporates continuous transcriptomic data with metabolic and regulatory networks via linear constraints linking genes and reactions, capturing transcriptional regulation. iMAT combines transcriptomic and proteomic data with genome-scale metabolic models using MILP optimization without requiring predefined objectives. E-Flux constrains reaction fluxes based on gene expression levels, allowing condition-specific modeling. E-Flux2 is inspired by E-flux and generates unique flux distributions with flexible objectives.
Results: The comparative analysis reveals differences in data integration strategies, computational requirements, and suitability for applications. GIMME has been successfully applied in microorganisms such as E. coli and in human skeletal muscle cells. TRFBA enhances prediction of cellular behavior under varying genetic and environmental conditions. iMAT delivers comprehensive metabolic states for diverse models, though it is computationally intensive. E-Flux enables rapid simulation of large-scale models, predicting responses to genetic, environmental, or pharmacological perturbations. E-Flux2 improves flux reproducibility and flexibility in objective functions, though validation has been mainly in central carbon metabolism.
Conclusion: This review provides a comparative overview of transcriptome-integrated FBA algorithms, summarizing methodologies, computational features, strengths, and limitations. It guides researchers in selecting appropriate algorithms for specific studies, facilitating accurate modeling of metabolism in diverse biological contexts and biomedical applications.