A comprehensive evaluation of multicategory classification methods for microbiomic data

Alexander Statnikov; Mikael Henaff; Varun Narendra; Kranti Konganti; Zhiguo Li; Liying Yang; Zhiheng Pei; Martin J. Blaser; Constantin F. Aliferis; Alexander V. Alekseyenko

doi:10.1186/2049-2618-1-11

A comprehensive evaluation of multicategory classification methods for microbiomic data

Alexander Statnikov, Mikael Henaff, Varun Narendra, Kranti Konganti, Zhiguo Li, Liying Yang, Zhiheng Pei, Martin J. Blaser, Constantin F. Aliferis, Alexander V. Alekseyenko

Institute for Health Informatics

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

149 Scopus citations

Abstract

Background: Recent advances in next-generation DNA sequencing enable rapid high-throughput quantitation of microbial community composition in human samples, opening up a new field of microbiomics. One of the promises of this field is linking abundances of microbial taxa to phenotypic and physiological states, which can inform development of new diagnostic, personalized medicine, and forensic modalities. Prior research has demonstrated the feasibility of applying machine learning methods to perform body site and subject classification with microbiomic data. However, it is currently unknown which classifiers perform best among the many available alternatives for classification with microbiomic data. Results: In this work, we performed a systematic comparison of 18 major classification methods, 5 feature selection methods, and 2 accuracy metrics using 8 datasets spanning 1,802 human samples and various classification tasks: body site and subject classification and diagnosis. Conclusions: We found that random forests, support vector machines, kernel ridge regression, and Bayesian logistic regression with Laplace priors are the most effective machine learning techniques for performing accurate classification from these microbiomic data.

Original language	English (US)
Article number	11
Journal	Microbiome
Volume	1
Issue number	1
DOIs	https://doi.org/10.1186/2049-2618-1-11
State	Published - Apr 5 2013

Bibliographical note

Funding Information:
This research was supported in part by grants UH2 AR057506-01S1 and UH3 CA140233 from the Human Microbiome Project and 1UL1 RR029893 from the National Center for Research Resources, National Institutes of Health, by the Diane Belfer Program in Human Microbial Ecology, and by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development. The authors acknowledge Efstratios Efstathiadis and Eric Peskin for providing access and support with high performance computing and Yingfei Ma for contribution to preparation of the PDX and PBS datasets. The authors also are grateful to Dan Knights and Rob Knight for contributing data from their study [2] and providing the technical details for reproducing their findings.

Keywords

Classification
Feature selection
Machine learning
Microbiomic data

Publisher link

10.1186/2049-2618-1-11

Find It

Find It at U of M Libraries

Cite this

@article{a820f7153e0242f7ad101722b61850a9,

title = "A comprehensive evaluation of multicategory classification methods for microbiomic data",

abstract = "Background: Recent advances in next-generation DNA sequencing enable rapid high-throughput quantitation of microbial community composition in human samples, opening up a new field of microbiomics. One of the promises of this field is linking abundances of microbial taxa to phenotypic and physiological states, which can inform development of new diagnostic, personalized medicine, and forensic modalities. Prior research has demonstrated the feasibility of applying machine learning methods to perform body site and subject classification with microbiomic data. However, it is currently unknown which classifiers perform best among the many available alternatives for classification with microbiomic data. Results: In this work, we performed a systematic comparison of 18 major classification methods, 5 feature selection methods, and 2 accuracy metrics using 8 datasets spanning 1,802 human samples and various classification tasks: body site and subject classification and diagnosis. Conclusions: We found that random forests, support vector machines, kernel ridge regression, and Bayesian logistic regression with Laplace priors are the most effective machine learning techniques for performing accurate classification from these microbiomic data.",

keywords = "Classification, Feature selection, Machine learning, Microbiomic data",

author = "Alexander Statnikov and Mikael Henaff and Varun Narendra and Kranti Konganti and Zhiguo Li and Liying Yang and Zhiheng Pei and Blaser, {Martin J.} and Aliferis, {Constantin F.} and Alekseyenko, {Alexander V.}",

note = "Funding Information: This research was supported in part by grants UH2 AR057506-01S1 and UH3 CA140233 from the Human Microbiome Project and 1UL1 RR029893 from the National Center for Research Resources, National Institutes of Health, by the Diane Belfer Program in Human Microbial Ecology, and by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development. The authors acknowledge Efstratios Efstathiadis and Eric Peskin for providing access and support with high performance computing and Yingfei Ma for contribution to preparation of the PDX and PBS datasets. The authors also are grateful to Dan Knights and Rob Knight for contributing data from their study [2] and providing the technical details for reproducing their findings.",

year = "2013",

month = apr,

day = "5",

doi = "10.1186/2049-2618-1-11",

language = "English (US)",

volume = "1",

journal = "Microbiome",

issn = "2049-2618",

publisher = "BioMed Central Ltd.",

number = "1",

}

TY - JOUR

T1 - A comprehensive evaluation of multicategory classification methods for microbiomic data

AU - Statnikov, Alexander

AU - Henaff, Mikael

AU - Narendra, Varun

AU - Konganti, Kranti

AU - Li, Zhiguo

AU - Yang, Liying

AU - Pei, Zhiheng

AU - Blaser, Martin J.

AU - Aliferis, Constantin F.

AU - Alekseyenko, Alexander V.

N1 - Funding Information: This research was supported in part by grants UH2 AR057506-01S1 and UH3 CA140233 from the Human Microbiome Project and 1UL1 RR029893 from the National Center for Research Resources, National Institutes of Health, by the Diane Belfer Program in Human Microbial Ecology, and by the Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development. The authors acknowledge Efstratios Efstathiadis and Eric Peskin for providing access and support with high performance computing and Yingfei Ma for contribution to preparation of the PDX and PBS datasets. The authors also are grateful to Dan Knights and Rob Knight for contributing data from their study [2] and providing the technical details for reproducing their findings.

PY - 2013/4/5

Y1 - 2013/4/5

N2 - Background: Recent advances in next-generation DNA sequencing enable rapid high-throughput quantitation of microbial community composition in human samples, opening up a new field of microbiomics. One of the promises of this field is linking abundances of microbial taxa to phenotypic and physiological states, which can inform development of new diagnostic, personalized medicine, and forensic modalities. Prior research has demonstrated the feasibility of applying machine learning methods to perform body site and subject classification with microbiomic data. However, it is currently unknown which classifiers perform best among the many available alternatives for classification with microbiomic data. Results: In this work, we performed a systematic comparison of 18 major classification methods, 5 feature selection methods, and 2 accuracy metrics using 8 datasets spanning 1,802 human samples and various classification tasks: body site and subject classification and diagnosis. Conclusions: We found that random forests, support vector machines, kernel ridge regression, and Bayesian logistic regression with Laplace priors are the most effective machine learning techniques for performing accurate classification from these microbiomic data.

AB - Background: Recent advances in next-generation DNA sequencing enable rapid high-throughput quantitation of microbial community composition in human samples, opening up a new field of microbiomics. One of the promises of this field is linking abundances of microbial taxa to phenotypic and physiological states, which can inform development of new diagnostic, personalized medicine, and forensic modalities. Prior research has demonstrated the feasibility of applying machine learning methods to perform body site and subject classification with microbiomic data. However, it is currently unknown which classifiers perform best among the many available alternatives for classification with microbiomic data. Results: In this work, we performed a systematic comparison of 18 major classification methods, 5 feature selection methods, and 2 accuracy metrics using 8 datasets spanning 1,802 human samples and various classification tasks: body site and subject classification and diagnosis. Conclusions: We found that random forests, support vector machines, kernel ridge regression, and Bayesian logistic regression with Laplace priors are the most effective machine learning techniques for performing accurate classification from these microbiomic data.

KW - Classification

KW - Feature selection

KW - Machine learning

KW - Microbiomic data

UR - http://www.scopus.com/inward/record.url?scp=84923857557&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84923857557&partnerID=8YFLogxK

U2 - 10.1186/2049-2618-1-11

DO - 10.1186/2049-2618-1-11

M3 - Article

AN - SCOPUS:84923857557

SN - 2049-2618

VL - 1

JO - Microbiome

JF - Microbiome

IS - 1

M1 - 11

ER -

A comprehensive evaluation of multicategory classification methods for microbiomic data

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this