The era of big data: Genome-scale modelling meets machine learning

Area of Science:

• Systems biology
• Computational biology
• Bioinformatics

Background:

• Omics data generation is rapidly increasing, necessitating advanced analytical tools.
• Genome-scale modeling (GSM) is crucial for organizing and analyzing this data.
• Machine learning (ML) offers complementary approaches, especially when biological mechanisms are unknown or for data pre-processing.

Purpose of the Study:

• To review recent advancements in genome-scale modeling and machine learning for omics data analysis.
• To explore the integration of these methodologies for a comprehensive understanding of biological systems.
• To highlight applications in microbial and mammalian systems.

Main Methods:

• Discussion of latest advances in genome-scale modeling techniques.
• Review of optimization algorithms for network and error reduction in GSM.
• Exploration of supervised and unsupervised machine learning applications on omics datasets.
• Presentation of hybrid modeling strategies combining GSM and ML.

Main Results:

• Genome-scale modeling is enhanced by new optimization algorithms for network refinement and strain design.
• Machine learning methods effectively analyze diverse omics data from microbial and mammalian cells.
• Hybrid models show promise in leveraging the strengths of both GSM and ML.

Conclusions:

• Integrating genome-scale modeling and machine learning provides powerful tools for omics data interpretation.
• These combined approaches facilitate deeper insights into cellular mechanisms and enable predictive applications.
• Future research should focus on further developing and applying hybrid modeling strategies.