aiSysMet: AI-powered systems metabolomics for biomarker discovery

Area of Science:

• Systems biology
• Metabolomics
• Multi-omics integration

Background:

• Metabolomics is crucial for understanding disease mechanisms and is often performed using liquid chromatography-mass spectrometry (LC-MS).
• Untargeted LC-MS studies face challenges with unannotated metabolites due to incomplete spectral libraries and data processing issues.
• These limitations impede the integration of metabolomics data with other omics layers.

Purpose of the Study:

• To develop an advanced platform for processing metabolomics data, annotating metabolites, and integrating multi-omics information.
• To overcome the limitations of current untargeted LC-MS data analysis.

Main Methods:

• Development of an AI-powered platform (aiSysMet).
• Utilizes statistical, machine learning, and deep learning algorithms.
• Features an interactive, modular web interface for cloud-based data analysis pipeline construction.

Main Results:

• aiSysMet provides robust metabolomics data processing and metabolite annotation.
• The platform facilitates the integrative analysis of multi-omics data.
• Enables users to build and execute custom data analysis pipelines in the cloud.

Conclusions:

• aiSysMet offers a comprehensive solution for enhancing metabolomics data analysis and multi-omics integration.
• The platform addresses key challenges in metabolite annotation and data processing.
• aiSysMet is accessible for non-commercial use, promoting wider adoption in research.