Biomarkers are used to predict quantitative metabolite concentration profiles in human red blood cells

PLoS Comput Biol. 2017 Mar 6;13(3):e1005424. doi: 10.1371/journal.pcbi.1005424. eCollection 2017 Mar.

Abstract

Deep-coverage metabolomic profiling has revealed a well-defined development of metabolic decay in human red blood cells (RBCs) under cold storage conditions. A set of extracellular biomarkers has been recently identified that reliably defines the qualitative state of the metabolic network throughout this metabolic decay process. Here, we extend the utility of these biomarkers by using them to quantitatively predict the concentrations of other metabolites in the red blood cell. We are able to accurately predict the concentration profile of 84 of the 91 (92%) measured metabolites (p < 0.05) in RBC metabolism using only measurements of these five biomarkers. The median of prediction errors (symmetric mean absolute percent error) across all metabolites was 13%. The ability to predict numerous metabolite concentrations from a simple set of biomarkers offers the potential for the development of a powerful workflow that could be used to evaluate the metabolic state of a biological system using a minimal set of measurements.

Publication types

  • Research Support, U.S. Gov't, Non-P.H.S.
  • Research Support, N.I.H., Extramural

MeSH terms

  • Biomarkers / blood*
  • Blood Proteins / metabolism*
  • Cells, Cultured
  • Computer Simulation
  • Erythrocytes / metabolism*
  • High-Throughput Screening Assays / methods*
  • Humans
  • Metabolic Flux Analysis / methods*
  • Metabolome / physiology*
  • Models, Cardiovascular
  • Reproducibility of Results
  • Sensitivity and Specificity

Substances

  • Biomarkers
  • Blood Proteins