Validation of metabolic pathway databases based on chemical substructure search

Area of Science:

• Biochemistry
• Bioinformatics
• Computational Biology

Background:

• Metabolic pathway databases, such as KEGG, house extensive biochemical reaction data.
• Maintaining the consistency of these large-scale databases is crucial for accurate biomedical research.

Purpose of the Study:

• To present a new computational method for assessing the consistency of biochemical reactions.
• To address the challenge of ensuring data integrity in large metabolic pathway databases.

Main Methods:

• Developed a novel method leveraging atomic rearrangement patterns in biochemical reactions.
• Reduced the atom mapping problem to chemical substructure searches between substrates and products.
• Applied the method to validate a significant portion of the KEGG LIGAND database.

Main Results:

• Successfully validated a substantial part of the KEGG LIGAND database using the new method.
• Demonstrated the efficacy of the atomic rearrangement pattern approach for consistency checking.
• The method effectively simplifies atom mapping for biochemical reactions.

Conclusions:

• The proposed method provides a robust way to determine the consistency of biochemical reactions.
• This contributes to improving the reliability and usability of metabolic pathway databases.
• The findings support enhanced data quality in bioinformatics and computational biology.