Learning representations of microbe–metabolite interactions

James T. Morton; Alexander A. Aksenov; Louis Felix Nothias; James R. Foulds; Robert A. Quinn; Michelle H. Badri; Tami L. Swenson; Marc W. Van Goethem; Trent R. Northen; Yoshiki Vazquez-Baeza; Mingxun Wang; Nicholas A. Bokulich; Aaron Watters; Se Jin Song; Richard Bonneau; Pieter C. Dorrestein; Rob Knight

doi:10.1038/s41592-019-0616-3

Learning representations of microbe–metabolite interactions

James T. Morton, Alexander A. Aksenov, Louis Felix Nothias, James R. Foulds, Robert A. Quinn, Michelle H. Badri, Tami L. Swenson, Marc W. Van Goethem, Trent R. Northen, Yoshiki Vazquez-Baeza, Mingxun Wang, Nicholas A. Bokulich, Aaron Watters, Se Jin Song, Richard Bonneau, Pieter C. Dorrestein, Rob Knight

Research output: Contribution to journal › Article › peer-review

Abstract

Integrating multiomics datasets is critical for microbiome research; however, inferring interactions across omics datasets has multiple statistical challenges. We solve this problem by using neural networks (https://github.com/biocore/mmvec) to estimate the conditional probability that each molecule is present given the presence of a specific microorganism. We show with known environmental (desert soil biocrust wetting) and clinical (cystic fibrosis lung) examples, our ability to recover microbe–metabolite relationships, and demonstrate how the method can discover relationships between microbially produced metabolites and inflammatory bowel disease.

Original language	English (US)
Pages (from-to)	1306-1314
Number of pages	9
Journal	Nature methods
Volume	16
Issue number	12
DOIs	https://doi.org/10.1038/s41592-019-0616-3
State	Published - Dec 1 2019

ASJC Scopus subject areas

Biotechnology
Biochemistry
Molecular Biology
Cell Biology

Access to Document

10.1038/s41592-019-0616-3

Cite this

Morton, J. T., Aksenov, A. A., Nothias, L. F., Foulds, J. R., Quinn, R. A., Badri, M. H., Swenson, T. L., Van Goethem, M. W., Northen, T. R., Vazquez-Baeza, Y., Wang, M., Bokulich, N. A., Watters, A., Song, S. J., Bonneau, R., Dorrestein, P. C., & Knight, R. (2019). Learning representations of microbe–metabolite interactions. Nature methods, 16(12), 1306-1314. https://doi.org/10.1038/s41592-019-0616-3

Morton, JT, Aksenov, AA, Nothias, LF, Foulds, JR, Quinn, RA, Badri, MH, Swenson, TL, Van Goethem, MW, Northen, TR, Vazquez-Baeza, Y, Wang, M, Bokulich, NA, Watters, A, Song, SJ, Bonneau, R, Dorrestein, PC & Knight, R 2019, 'Learning representations of microbe–metabolite interactions', Nature methods, vol. 16, no. 12, pp. 1306-1314. https://doi.org/10.1038/s41592-019-0616-3

@article{033d42cc619f494db4e5ab72bcc5ffab,

title = "Learning representations of microbe–metabolite interactions",

abstract = "Integrating multiomics datasets is critical for microbiome research; however, inferring interactions across omics datasets has multiple statistical challenges. We solve this problem by using neural networks (https://github.com/biocore/mmvec) to estimate the conditional probability that each molecule is present given the presence of a specific microorganism. We show with known environmental (desert soil biocrust wetting) and clinical (cystic fibrosis lung) examples, our ability to recover microbe–metabolite relationships, and demonstrate how the method can discover relationships between microbially produced metabolites and inflammatory bowel disease.",

author = "Morton, {James T.} and Aksenov, {Alexander A.} and Nothias, {Louis Felix} and Foulds, {James R.} and Quinn, {Robert A.} and Badri, {Michelle H.} and Swenson, {Tami L.} and {Van Goethem}, {Marc W.} and Northen, {Trent R.} and Yoshiki Vazquez-Baeza and Mingxun Wang and Bokulich, {Nicholas A.} and Aaron Watters and Song, {Se Jin} and Richard Bonneau and Dorrestein, {Pieter C.} and Rob Knight",

note = "Funding Information: We would like to thank V. Pawlowsky, J. J. Egozcue and S. Holmes for their insights on the geometry of this neural network model. In addition, we would also like to thank N. Bokulich for feedback and contributions on the mmvec software package. T.L.S., M.W.V.G. and T.R.N. acknowledge funding from the Office of Science Early Career Research Program, Office of Biological and Environmental Research of the U.S. Department of Energy under contract number DE-AC02-05CH11231 to Lawrence Berkeley National Laboratory. This study was in part supported by grant P41GM103484 for the Center for Computational Mass Spectrometry and instrument support through National Institutes of Health grants S10RR029121 and R03 CA211211 on reuse of metabolomics data. Y.V.B. is funded by the Janssen Human Microbiome Institute through a collaboration with the Center for Microbiome Innovation. J.T.M. was funded by National Science Foundation grant GRFP DGE-1144086. R.K. and S.J.S. have been funded by Janssen under grant number 20175015 and the Alfred P. Sloan Foundation under grant number G-2017-9838. Publisher Copyright: {\textcopyright} 2019, The Author(s), under exclusive licence to Springer Nature America, Inc.",

year = "2019",

month = dec,

day = "1",

doi = "10.1038/s41592-019-0616-3",

language = "English (US)",

volume = "16",

pages = "1306--1314",

journal = "Nature Methods",

issn = "1548-7091",

publisher = "Public Library of Science",

number = "12",

}

TY - JOUR

T1 - Learning representations of microbe–metabolite interactions

AU - Morton, James T.

AU - Aksenov, Alexander A.

AU - Nothias, Louis Felix

AU - Foulds, James R.

AU - Quinn, Robert A.

AU - Badri, Michelle H.

AU - Swenson, Tami L.

AU - Van Goethem, Marc W.

AU - Northen, Trent R.

AU - Vazquez-Baeza, Yoshiki

AU - Wang, Mingxun

AU - Bokulich, Nicholas A.

AU - Watters, Aaron

AU - Song, Se Jin

AU - Bonneau, Richard

AU - Dorrestein, Pieter C.

AU - Knight, Rob

N1 - Funding Information: We would like to thank V. Pawlowsky, J. J. Egozcue and S. Holmes for their insights on the geometry of this neural network model. In addition, we would also like to thank N. Bokulich for feedback and contributions on the mmvec software package. T.L.S., M.W.V.G. and T.R.N. acknowledge funding from the Office of Science Early Career Research Program, Office of Biological and Environmental Research of the U.S. Department of Energy under contract number DE-AC02-05CH11231 to Lawrence Berkeley National Laboratory. This study was in part supported by grant P41GM103484 for the Center for Computational Mass Spectrometry and instrument support through National Institutes of Health grants S10RR029121 and R03 CA211211 on reuse of metabolomics data. Y.V.B. is funded by the Janssen Human Microbiome Institute through a collaboration with the Center for Microbiome Innovation. J.T.M. was funded by National Science Foundation grant GRFP DGE-1144086. R.K. and S.J.S. have been funded by Janssen under grant number 20175015 and the Alfred P. Sloan Foundation under grant number G-2017-9838. Publisher Copyright: © 2019, The Author(s), under exclusive licence to Springer Nature America, Inc.

PY - 2019/12/1

Y1 - 2019/12/1

N2 - Integrating multiomics datasets is critical for microbiome research; however, inferring interactions across omics datasets has multiple statistical challenges. We solve this problem by using neural networks (https://github.com/biocore/mmvec) to estimate the conditional probability that each molecule is present given the presence of a specific microorganism. We show with known environmental (desert soil biocrust wetting) and clinical (cystic fibrosis lung) examples, our ability to recover microbe–metabolite relationships, and demonstrate how the method can discover relationships between microbially produced metabolites and inflammatory bowel disease.

AB - Integrating multiomics datasets is critical for microbiome research; however, inferring interactions across omics datasets has multiple statistical challenges. We solve this problem by using neural networks (https://github.com/biocore/mmvec) to estimate the conditional probability that each molecule is present given the presence of a specific microorganism. We show with known environmental (desert soil biocrust wetting) and clinical (cystic fibrosis lung) examples, our ability to recover microbe–metabolite relationships, and demonstrate how the method can discover relationships between microbially produced metabolites and inflammatory bowel disease.

UR - http://www.scopus.com/inward/record.url?scp=85075724305&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075724305&partnerID=8YFLogxK

U2 - 10.1038/s41592-019-0616-3

DO - 10.1038/s41592-019-0616-3

M3 - Article

C2 - 31686038

AN - SCOPUS:85075724305

SN - 1548-7091

VL - 16

SP - 1306

EP - 1314

JO - Nature Methods

JF - Nature Methods

IS - 12

ER -

Learning representations of microbe–metabolite interactions

Abstract

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this