Atom Identifiers Generated by a Neighborhood-Specific Graph Coloring Method Enable Compound Harmonization across Metabolic Databases

Benchmark dataset for training machine learning models to predict the pathway involvement of metabolites

Metabolic pathways are a human-defined grouping of life sustaining biochemical reactions, metabolites being both the reactants and products of these reactions. But many public datasets include identified metabolites whose pathway involvement is unknown, hindering metabolic interpretation. To address these shortcomings, various machine learning models, including those trained on data from the Kyoto Encyclopedia of Genes and Genomes (KEGG), have been developed to predict the pathway involvement of metabolites based on their chemical descriptions; however, these prior models are based on old metabolite KEGG-based datasets, including one benchmark dataset that is invalid due to the presence of over 1500 duplicate entries. Therefore, we have developed a new benchmark dataset derived from the KEGG following optimal standards of scientific computational reproducibility and including all source code needed to update the benchmark dataset as KEGG changes. We have used this new benchmark dataset with our atom coloring methodology to develop and compare the performance of Random Forest, XGBoost, and multilayer perceptron with autoencoder models generated from our new benchmark dataset. Best overall weighted average performance across 1000 unique folds was an F1-score of 0.8180 and Matthews correlation coefficient of 0.7933, which was provided by XGBoost binary classification models for 11 KEGG-defined pathway categories.

Isotope-Assisted Metabolic Flux Analysis: A Powerful Technique to Gain New Insights into the Human Metabolome in Health and Disease

Cell metabolism represents the coordinated changes in genes, proteins, and metabolites that occur in health and disease. The metabolic fluxome, which includes both intracellular and extracellular metabolic reaction rates (fluxes), therefore provides a powerful, integrated description of cellular phenotype. However, intracellular fluxes cannot be directly measured. Instead, flux quantification requires sophisticated mathematical and computational analysis of data from isotope labeling experiments. In this review, we describe isotope-assisted metabolic flux analysis (iMFA), a rigorous computational approach to fluxome quantification that integrates metabolic network models and experimental data to generate quantitative metabolic flux maps. We highlight practical considerations for implementing iMFA in mammalian models, as well as iMFA applications in in vitro and in vivo studies of physiology and disease. Finally, we identify promising new frontiers in iMFA which may enable us to fully unlock the potential of iMFA in biomedical research.

Hierarchical Harmonization of Atom-Resolved Metabolic Reactions across Metabolic Databases

Metabolic models have been proven to be useful tools in system biology and have been successfully applied to various research fields in a wide range of organisms. A relatively complete metabolic network is a prerequisite for deriving reliable metabolic models. The first step in constructing metabolic network is to harmonize compounds and reactions across different metabolic databases. However, effectively integrating data from various sources still remains a big challenge. Incomplete and inconsistent atomistic details in compound representations across databases is a very important limiting factor. Here, we optimized a subgraph isomorphism detection algorithm to validate generic compound pairs. Moreover, we defined a set of harmonization relationship types between compounds to deal with inconsistent chemical details while successfully capturing atom-level characteristics, enabling a more complete enabling compound harmonization across metabolic databases. In total, 15,704 compound pairs across KEGG (Kyoto Encyclopedia of Genes and Genomes) and MetaCyc databases were detected. Furthermore, utilizing the classification of compound pairs and EC (Enzyme Commission) numbers of reactions, we established hierarchical relationships between metabolic reactions, enabling the harmonization of 3856 reaction pairs. In addition, we created and used atom-specific identifiers to evaluate the consistency of atom mappings within and between harmonized reactions, detecting some consistency issues between the reaction and compound descriptions in these metabolic databases.

Analysis of steady-state carbon tracer experiments using akaike information criteria

INTRODUCTION

Carbon isotope tracers have been used to determine relative rates of tricarboxylic acid cycle (TCA) cycle pathways since the 1950s. Steady-state experimental data are typically fit to a single mathematical model of metabolism to determine metabolic fluxes. Whether the chosen model is appropriate for the biological system has generally not been evaluated systematically. An overly-simple model omits known pathways while an overly-complex model may produce incorrect results due to overfitting.

OBJECTIVES

The objectives were to develop and study a method that systematically evaluates multiple TCA cycle mathematical models as part of the fitting process.

METHODS

The problem of choosing overly-simple or overly-complex models was approached by developing software that automatically explores all possible combinations of flux through pyruvate dehydrogenase, pyruvate kinase, pyruvate carboxylase and anaplerosis at propionyl-CoA carboxylase, and equivalent pathways, all relative to TCA cycle flux. Typical TCA cycle metabolic tracer experiments that use 13C nuclear magnetic resonance for detection and quantification of 13C-enriched glutamate products were simulated and analyzed. By evaluating the multiple model fits with both the conventional sum-of-squares residual error (SSRE) and the Akaike Information Criterion (AIC), the software helps the investigator understand the interaction between model complexity and goodness of fit.

RESULTS

When fitting alternative models of the TCA cycle metabolism, the SSRE may identify more than one model that fits the data well. Among those models, the AIC provides guidance as to which is the simplest of the candidate models is sufficient to describe the observed data. However under some conditions, AIC used alone inappropriately discriminates against necessary metabolic complexity.

CONCLUSION

In combination, the SSRE and AIC help the investigator identify the model that best describes the metabolism of a biological system.