Machine learning approaches in microbiome research: challenges and best practices
Microbiome data predictive analysis within a machine learning (ML) workflow presents numerous domain-specific challenges involving preprocessing, feature selection, predictive modeling, performance estimation, model interpretation, and the extraction of biological information from the results. To as...
| Main Authors: | , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-09-01
|
| Series: | Frontiers in Microbiology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmicb.2023.1261889/full |
| _version_ | 1797676893775855616 |
|---|---|
| author | Georgios Papoutsoglou Georgios Papoutsoglou Sonia Tarazona Marta B. Lopes Marta B. Lopes Thomas Klammsteiner Thomas Klammsteiner Eliana Ibrahimi Julia Eckenberger Julia Eckenberger Pierfrancesco Novielli Pierfrancesco Novielli Alberto Tonda Alberto Tonda Andrea Simeon Rajesh Shigdel Stéphane Béreux Stéphane Béreux Giacomo Vitali Sabina Tangaro Sabina Tangaro Leo Lahti Andriy Temko Marcus J. Claesson Marcus J. Claesson Magali Berland |
| author_facet | Georgios Papoutsoglou Georgios Papoutsoglou Sonia Tarazona Marta B. Lopes Marta B. Lopes Thomas Klammsteiner Thomas Klammsteiner Eliana Ibrahimi Julia Eckenberger Julia Eckenberger Pierfrancesco Novielli Pierfrancesco Novielli Alberto Tonda Alberto Tonda Andrea Simeon Rajesh Shigdel Stéphane Béreux Stéphane Béreux Giacomo Vitali Sabina Tangaro Sabina Tangaro Leo Lahti Andriy Temko Marcus J. Claesson Marcus J. Claesson Magali Berland |
| author_sort | Georgios Papoutsoglou |
| collection | DOAJ |
| description | Microbiome data predictive analysis within a machine learning (ML) workflow presents numerous domain-specific challenges involving preprocessing, feature selection, predictive modeling, performance estimation, model interpretation, and the extraction of biological information from the results. To assist decision-making, we offer a set of recommendations on algorithm selection, pipeline creation and evaluation, stemming from the COST Action ML4Microbiome. We compared the suggested approaches on a multi-cohort shotgun metagenomics dataset of colorectal cancer patients, focusing on their performance in disease diagnosis and biomarker discovery. It is demonstrated that the use of compositional transformations and filtering methods as part of data preprocessing does not always improve the predictive performance of a model. In contrast, the multivariate feature selection, such as the Statistically Equivalent Signatures algorithm, was effective in reducing the classification error. When validated on a separate test dataset, this algorithm in combination with random forest modeling, provided the most accurate performance estimates. Lastly, we showed how linear modeling by logistic regression coupled with visualization techniques such as Individual Conditional Expectation (ICE) plots can yield interpretable results and offer biological insights. These findings are significant for clinicians and non-experts alike in translational applications. |
| first_indexed | 2024-03-11T22:37:05Z |
| format | Article |
| id | doaj.art-a35769259f1946d3b9af729120cec4fd |
| institution | Directory Open Access Journal |
| issn | 1664-302X |
| language | English |
| last_indexed | 2024-03-11T22:37:05Z |
| publishDate | 2023-09-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Microbiology |
| spelling | doaj.art-a35769259f1946d3b9af729120cec4fd2023-09-22T13:10:54ZengFrontiers Media S.A.Frontiers in Microbiology1664-302X2023-09-011410.3389/fmicb.2023.12618891261889Machine learning approaches in microbiome research: challenges and best practicesGeorgios Papoutsoglou0Georgios Papoutsoglou1Sonia Tarazona2Marta B. Lopes3Marta B. Lopes4Thomas Klammsteiner5Thomas Klammsteiner6Eliana Ibrahimi7Julia Eckenberger8Julia Eckenberger9Pierfrancesco Novielli10Pierfrancesco Novielli11Alberto Tonda12Alberto Tonda13Andrea Simeon14Rajesh Shigdel15Stéphane Béreux16Stéphane Béreux17Giacomo Vitali18Sabina Tangaro19Sabina Tangaro20Leo Lahti21Andriy Temko22Marcus J. Claesson23Marcus J. Claesson24Magali Berland25Department of Computer Science, University of Crete, Heraklion, GreeceJADBio Gnosis DA S.A., Science and Technology Park of Crete, Heraklion, GreeceDepartment of Applied Statistics and Operations Research and Quality, Polytechnic University of Valencia, Valencia, SpainCenter for Mathematics and Applications (NOVA Math), NOVA School of Science and Technology, Caparica, PortugalResearch and Development Unit for Mechanical and Industrial Engineering (UNIDEMI), Department of Mechanical and Industrial Engineering, NOVA School of Science and Technology, Caparica, PortugalDepartment of Ecology, Universität Innsbruck, Innsbruck, AustriaDepartment of Microbiology, Universität Innsbruck, Innsbruck, AustriaDepartment of Biology, University of Tirana, Tirana, AlbaniaSchool of Microbiology, University College Cork, Cork, Ireland0APC Microbiome Ireland, Cork, Ireland1Department of Soil, Plant, and Food Sciences, University of Bari Aldo Moro, Bari, Italy2National Institute for Nuclear Physics, Bari Division, Bari, Italy3UMR 518 MIA-PS, INRAE, Paris-Saclay University, Palaiseau, France4Complex Systems Institute of Paris Ile-de-France (ISC-PIF) - UAR 3611 CNRS, Paris, France5BioSense Institute, University of Novi Sad, Novi Sad, Serbia6Department of Clinical Science, University of Bergen, Bergen, Norway7MetaGenoPolis, INRAE, Paris-Saclay University, Jouy-en-Josas, France8MaIAGE, INRAE, Paris-Saclay University, Jouy-en-Josas, France7MetaGenoPolis, INRAE, Paris-Saclay University, Jouy-en-Josas, France1Department of Soil, Plant, and Food Sciences, University of Bari Aldo Moro, Bari, Italy2National Institute for Nuclear Physics, Bari Division, Bari, Italy9Department of Computing, University of Turku, Turku, Finland0Department of Electrical and Electronic Engineering, University College Cork, Cork, IrelandSchool of Microbiology, University College Cork, Cork, Ireland0APC Microbiome Ireland, Cork, Ireland7MetaGenoPolis, INRAE, Paris-Saclay University, Jouy-en-Josas, FranceMicrobiome data predictive analysis within a machine learning (ML) workflow presents numerous domain-specific challenges involving preprocessing, feature selection, predictive modeling, performance estimation, model interpretation, and the extraction of biological information from the results. To assist decision-making, we offer a set of recommendations on algorithm selection, pipeline creation and evaluation, stemming from the COST Action ML4Microbiome. We compared the suggested approaches on a multi-cohort shotgun metagenomics dataset of colorectal cancer patients, focusing on their performance in disease diagnosis and biomarker discovery. It is demonstrated that the use of compositional transformations and filtering methods as part of data preprocessing does not always improve the predictive performance of a model. In contrast, the multivariate feature selection, such as the Statistically Equivalent Signatures algorithm, was effective in reducing the classification error. When validated on a separate test dataset, this algorithm in combination with random forest modeling, provided the most accurate performance estimates. Lastly, we showed how linear modeling by logistic regression coupled with visualization techniques such as Individual Conditional Expectation (ICE) plots can yield interpretable results and offer biological insights. These findings are significant for clinicians and non-experts alike in translational applications.https://www.frontiersin.org/articles/10.3389/fmicb.2023.1261889/fullmicrobiome data analysismachine learning methodspreprocessingfeature selectionpredictive modelingmodel selection |
| spellingShingle | Georgios Papoutsoglou Georgios Papoutsoglou Sonia Tarazona Marta B. Lopes Marta B. Lopes Thomas Klammsteiner Thomas Klammsteiner Eliana Ibrahimi Julia Eckenberger Julia Eckenberger Pierfrancesco Novielli Pierfrancesco Novielli Alberto Tonda Alberto Tonda Andrea Simeon Rajesh Shigdel Stéphane Béreux Stéphane Béreux Giacomo Vitali Sabina Tangaro Sabina Tangaro Leo Lahti Andriy Temko Marcus J. Claesson Marcus J. Claesson Magali Berland Machine learning approaches in microbiome research: challenges and best practices Frontiers in Microbiology microbiome data analysis machine learning methods preprocessing feature selection predictive modeling model selection |
| title | Machine learning approaches in microbiome research: challenges and best practices |
| title_full | Machine learning approaches in microbiome research: challenges and best practices |
| title_fullStr | Machine learning approaches in microbiome research: challenges and best practices |
| title_full_unstemmed | Machine learning approaches in microbiome research: challenges and best practices |
| title_short | Machine learning approaches in microbiome research: challenges and best practices |
| title_sort | machine learning approaches in microbiome research challenges and best practices |
| topic | microbiome data analysis machine learning methods preprocessing feature selection predictive modeling model selection |
| url | https://www.frontiersin.org/articles/10.3389/fmicb.2023.1261889/full |
| work_keys_str_mv | AT georgiospapoutsoglou machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT georgiospapoutsoglou machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT soniatarazona machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT martablopes machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT martablopes machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT thomasklammsteiner machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT thomasklammsteiner machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT elianaibrahimi machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT juliaeckenberger machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT juliaeckenberger machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT pierfrancesconovielli machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT pierfrancesconovielli machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT albertotonda machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT albertotonda machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT andreasimeon machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT rajeshshigdel machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT stephanebereux machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT stephanebereux machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT giacomovitali machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT sabinatangaro machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT sabinatangaro machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT leolahti machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT andriytemko machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT marcusjclaesson machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT marcusjclaesson machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices AT magaliberland machinelearningapproachesinmicrobiomeresearchchallengesandbestpractices |