Mapping human metabolic pathways in the small molecule chemical space

Area of Science:

• Biochemistry
• Computational Chemistry
• Pharmacology

Background:

• Human metabolic pathways are complex systems involving numerous small molecules.
• Understanding the relationships between these pathways and molecules is crucial for drug discovery and development.
• Current methods for analyzing pathway-molecule interactions in chemical space are limited.

Purpose of the Study:

• To analyze the clustering and overlap of human metabolic pathways within chemical space.
• To develop and validate a statistical model for predicting small molecule proximity to metabolic pathways.
• To assess the proximity of marketed drugs to human metabolic pathways and predict potential cross-interactions.

Main Methods:

• Utilized visual and quantitative approaches to analyze metabolic pathway distribution and overlap in chemical space.
• Developed a classifier using physicochemical and topological descriptors to predict metabolic pathway membership for small molecules.
• Applied the developed model to evaluate marketed drugs' proximity to human metabolic pathways.

Main Results:

• Revealed the distribution, overlap, and relationships of human metabolic pathways using selected descriptors.
• The predictive model demonstrated good performance for isolated pathways but reduced accuracy for overlapping pathways.
• Examined drug overlap with the human metabolome and predicted potential cross-interactions with major metabolic pathways.

Conclusions:

• Human metabolic pathways exhibit varying degrees of clustering and overlap in chemical space.
• The developed predictive model offers insights into small molecule-pathway relationships, with performance dependent on pathway isolation.
• The study provides a framework for understanding drug interactions with the human metabolome, highlighting potential cross-interactions.