Huan Jin

md_harmonize harmonizes compounds and metabolic reactions across databases by generating unique identifiers based on atom structures for improved metabolic network analysis.

The paper develops a new benchmark dataset for predicting metabolite pathway involvement using machine learning models and evaluates their performance.

The paper harmonizes metabolic reactions and compounds across databases by improving atom-level consistency and establishing hierarchical relationships.

The paper introduces a graph coloring method to generate atom identifiers for compounds, enabling compound harmonization across metabolic databases and improving metabolic network integration.

The paper develops a software package for selecting metabolic models using isotopologue data from SIRM experiments, demonstrating its effectiveness with UDP-GlcNAc datasets.

The paper introduces a framework for deconvolving isotopologue profiles using moiety models and provides Python packages for optimization and model selection.