Accurate, fully-automated NMR spectral profiling for metabolomics
Many diseases cause significant changes to the concentrations of small molecules (aka metabolites) that appear in a person's biofluids,
which means such diseases can often be readily detected from a person’s “metabolic profile”.
This information can be extracted from a person's biofluids using NMR spectroscopy.
Today, this is often done manually by trained experts, which means this process is relatively slow, expensive and can be error-prone.
A system that can quickly, accurately and autonomously produce a person's metabolic profile would enable efficient and reliable prediction of many such diseases from a single sample,
which could significantly improve the way medicine is practiced.
This paper presents such a system:
Given a 1D
1H NMR pectrum of a complex biofluid such as serum or CSF,
our Bayesil system can automatically determine this metabolic profile, and do so without any human guidance.
This requires first performing all of the required spectral processing steps
(ie, Fourier transformation, phasing, solvent-removal, chemical shift referencing, baseline correction, lineshape convolution)
then matching this resulting spectrum against a reference compound library, which contains the “signatures” of each relevant metabolite.
Many of these processing steps are novel algorithms, and
our matching step views spectral matching as an inference problem within a probabilistic graphical model that rapidly approximates the most probable metabolic profile.
Our extensive studies on a diverse set of complex mixtures
(real biological samples, defined mixtures and realistic computer generated spectra; each involving ~50 compounds),
show that Bayesil can autonomously and accurately find NMR-detectable metabolites at concentrations as low as 2μM,
in terms of both identification (~90% correct) and quantification (~10% error),
in under 5 minutes on a single CPU processor.
These results demonstrate that Bayesil is the first fully-automatic publicly-accessible system that provides quantitative NMR spectral profiling effectively --
with an accuracy that meets or exceeds the performance of highly trained human experts.
We anticipate this tool will usher in high-throughput metabolomics and enable a wealth of new applications of NMR in clinical settings.
Users can access Bayesil at http://www.bayesil.ca .
Citation
S. Ravanbakhsh,
P. Liu,
T. Bjorndahl,
R. Mandal,
J. Grant,
M. Wilson,
R. Eisner,
I. Sinelnikov,
X. Hu,
C. Luchinat,
R. Greiner,
D. Wishart.
"Accurate, fully-automated NMR spectral profiling for metabolomics".
PLoS One, 10(5), pp e0132873, May 2015.
| Keywords: |
NMR, metabolomics, bioinformatics, spectroscopy, metabolites |
| Category: |
In Journal |
| Web Links: |
DOI |
| |
Journal |
BibTeX
@article{Ravanbakhsh+al:PLoSONE15,
author = {Siamak Ravanbakhsh and Philip Liu and Trent Bjorndahl and Rupasri
Mandal and Jason Grant and Michael Wilson and Roman Eisner and Igor
Sinelnikov and Xiaoyu Hu and Claudio Luchinat and Russ Greiner and David S.
Wishart},
title = {Accurate, fully-automated NMR spectral profiling for metabolomics},
Volume = "10",
Number = "5",
Pages = {e0132873},
journal = {PLoS One},
year = 2015,
}Last Updated: February 10, 2020
Submitted by Sabina P 
