Clinical diagnosis of metabolic disorders using untargeted metabolomic profiling and disease-specific networks learned from profiling data

Abstract

Untargeted metabolomics is a global molecular profiling technology that can be used to screen for inborn errors of metabolism (IEMs). Metabolite perturbations are evaluated based on current knowledge of specific metabolic pathway deficiencies, a manual diagnostic process that is qualitative, has limited scalability, and is not equipped to learn from accumulating clinical data. Our purpose was to improve upon manual diagnosis of IEMs in the clinic by developing novel computational methods for analyzing untargeted metabolomics data. We employed CTD, an automated computational diagnostic method that “connects the dots” between metabolite perturbations observed in individual metabolomics profiling data and modules identified in disease­specific metabolite co-perturbation networks learned from prior profiling data. We also extended CTD to calculate distances between any two individuals (CTDncd) and between an individual and a disease state (CTDdm), to provide additional network-quantified predictors for use in diagnosis. We show that across 539 plasma samples, CTD-based network-quantified measures can reproduce accurate diagnosis of 16 different IEMs, including adenylosuccinase deficiency, argininemia, argininosuccinic aciduria, aromatic l-amino acid decarboxylase deficiency, cerebral creatine deficiency syndrome type 2, citrullinemia, cobalamin biosynthesis defect, GABA-transaminase deficiency, glutaric acidemia type 1, maple syrup urine disease, methylmalonic aciduria, ornithine transcarbamylase deficiency, phenylketonuria, propionic acidemia, rhizomelic chondrodysplasia punctata, and the Zellweger spectrum disorders. Our approach can be used to supplement information from biochemical pathways and has the potential to significantly enhance the interpretation of variants of uncertain significance uncovered by exome sequencing. CTD, CTDdm, and CTDncd can serve as an essential toolset for biological interpretation of untargeted metabolomics data that overcomes limitations associated with manual diagnosis to assist diagnosticians in clinical decision-making. By automating and quantifying the interpretation of perturbation patterns, CTD can improve the speed and confidence by which clinical laboratory directors make diagnostic and treatment decisions, while automatically improving performance with new case data.

Circulating N-formylmethionine and metabolic shift in critical illness: a multicohort metabolomics study

“…Metabolomics can bridge the knowledge gap between clinical heterogeneity and severity of critical illness [ 49 ]. Finding consistent metabolic shift patterns in clinically heterogeneous cohorts has the potential to identify biologically meaningful phenotypes [ 50 ].…”

An Information-Theoretic Bound on p-Values for Detecting Communities Shared between Weighted Labeled Graphs

“…In order to see how the effectiveness of the proposed approach is impacted by other graph and node module features, we have generated random synthetic graphs defined by five parameters. The values of parameters were assigned to be similar to potential biological use cases—disease-specific metabolite co-perturbation networks contain between 300 and 1000 nodes [ 4 , 5 ], while gene co-expression networks consist of thousands of nodes [ 19 ], but the connectedness patterns remain significantly smaller, consisting of no more than 1% of network nodes. An overview of the parameters used for synthetic graph generation and their values is given in Table 1 .…”

“…Finally, edges in metabolite co-perturbation networks can have negative weights between nodes corresponding to a substrate and a product around a perturbed enzyme [ 5 ]. However, this negative weight still expresses connectedness.…”

“…Namely, gene co-expression networks and metabolic networks are weighted undirected labeled graphs in which patterns of connectedness are very important to detect and evaluate. Our previous work illustrates successful application of information-theoretic methods to both gene expression networks in breast cancer [ 4 ] as well as to metabolic networks that model metabolomic perturbations in human inborn errors of metabolism [ 5 , 6 ]. A vast array of applications can be envisioned in social network analysis.…”

A narrative review of metabolomics in the era of “-omics”: integration into clinical practice for inborn errors of metabolism

“…Future trends and integration of untargeted metabolomics to the bedside Untargeted metabolomic technologies allow for analysis of a comprehensive array of metabolites that would otherwise require individualized assays and sample preparations (43). This technology is also likely to offer substantial clinical benefit not achieved by targeted metabolomic testing, and may be further enhanced by developments in machine learning (44) and statistical modelling (45,46). The clinical potential of untargeted metabolomics in diagnosis of IEM has been investigated extensively (11,13,15,18,22,33,(47)(48)(49)(50)(51)(52)(53), and complements other "-omic" approaches in the diagnosis of IEM and other genetic disorders (16,42,(54)(55)(56)(57)(58).…”