Discriminative Machine Learning Analysis for Skin Microbiome: Observing Biomarkers in Patients with Seborrheic Dermatitis


  • H.E.C. van der Wall Method Development, Centre for Human Drug Research, The Netherlands
  • R.J. Doll Method Development, Centre for Human Drug Research, The Netherlands
  • G.J.P. van Westen Drug Discovery and Safety, Leiden Academic Centre for Drug Research, The Netherlands
  • T. Niemeyer-van der Kolk Method Development, Centre for Human Drug Research, The Netherlands
  • G. Feiss GL Feiss Consulting, LLC, Telford, Pennsylvania, United States
  • H. Pinckaers Department of Pathology, Radboud University Medical Center, The Netherlands
  • M.B.A. van Doorn Department of Dermatology, Erasmus Medical Center, The Netherlands
  • T. Nijsten Department of Dermatology, Erasmus Medical Center, The Netherlands
  • M.G.H. Sanders Department of Dermatology, Erasmus Medical Center, The Netherlands
  • A.F. Cohen Method Development, Centre for Human Drug Research, The Netherlands
  • J. Burggraaf Method Development, Centre for Human Drug Research, The Netherlands
  • R. Rissmann Method Development, Centre for Human Drug Research, The Netherlands
  • L.M. Pardo Department of Dermatology, Erasmus Medical Center, The Netherlands




Microbiome, Artificial intelligence, Machine learning, Seborrheic dermatitis, Data science


In recent years the skin microbiome has taken center stage as drug target and as disease biomarker. Computational analyses of microbiome sequencing data from patients with skin diseases, for example seborrheic dermatitis, can be performed to identify discriminative biomarkers in the microbiome profile. The aim of the present study was twofold, namely to employ machine learning to predict disease from the microbiome dataset, and to identify discriminative biomarkers in the microbiome of patients with seborrheic dermatitis versus healthy controls using machine learning techniques. The population consisted of 97 patients with seborrheic dermatitis and 763 healthy controls. Skin swabs were taken from naso-labial fold (lesional skin: n = 22; non-lesional skin: n = 75, controls: n = 763). Using an extra trees machine learning model, differences between the skin microbiome of patients with seborrheic dermatitis versus healthy controls were characterized. Subsequently, the most important microorganisms for discrimination were determined by feature analysis and SHapley Additive exPlanations (SHAP) values. The accuracy of the prediction models to discriminate between skin affected by seborrheic dermatitis and facial skin from healthy subjects was 77% and the ROC-AUC was 83%. Next to Cutibacterium and Staphylococcus, the most important organisms for discrimination had a relatively low occurrence. Our study showed that machine learning can be utilized to identify discriminating biomarkers in the microbiome skin. This indicates that machine learning can be of major importance in basic skin research, and in the discovery and development of new individualized therapies, involving the microbiome.


E.A. Grice, J.A. Segre, The skin microbiome, Nat. Rev. Microbiol. 9 (2011), 244–253. https://doi.org/10.1038/nrmicro2537

M. Garcia-Garcerà, K. Garcia-Etxebarria, M. Coscollà, A. Latorre, F. Calafell, A new method for extracting skin microbes allows metagenomic analysis of whole-deep skin, PLoS ONE. 8 (2013), e74914. https://doi.org/10.1371/journal.pone.0074914

P.L.J.M. Zeeuwen, M. Kleerebezem, H.M. Timmerman, J. Schalkwijk, Microbiome and skin diseases, Curr. Opin. Allergy Clin. Immunol. 13 (2013), 514–520. https://doi.org/10.1097/aci.0b013e328364ebeb

M. Leclercq, B. Vittrant, M.L. Martin-Magniette, M.P. Scott Boyer, O. Perin, A. Bergeron, et al., Large-scale automatic feature selection for biomarker discovery in high-dimensional OMICs data, Front. Genet. 10 (2019), 452. https://doi.org/10.3389/fgene.2019.00452

D. Zhang, D. Shen, The Alzheimer's Disease Neuroimaging Initiative, Multi-modal multi-task learning for joint prediction of multiple regression and classification variables in Alzheimer's disease, NeuroImage. 59 (2012), 895–907. https://doi.org/10.1016/j.neuroimage.2011.09.069

G. Deshpande, L.E. Libero, K.R. Sreenivasan, H.D. Deshpande, R.K. Kana, Identification of neural connectivity signatures of autism using machine learning, Front. Hum. Neurosci. 7 (2013), 670. https://doi.org/10.3389/fnhum.2013.00670

T. Fekete, N. Zach, L.R. Mujica-Parodi, M.R. Turner, Multiple kernel learning captures a systems-level functional connectivity biomarker signature in amyotrophic lateral sclerosis, PLoS ONE. 8 (2013), e85190. https://doi.org/10.1371/journal.pone.0085190

A. Sutherland, M. Thomas, R.A. Brandon, R.B. Brandon, J. Lipman, B. Tang, et al., Development and validation of a novel molecular biomarker diagnostic test for the early detection of sepsis, Crit. Care. 15 (2011), R149. https://pubmed.ncbi.nlm.nih.gov/21682927/

A. Kohli, E.A. Holzwanger, A.N. Levy, Emerging use of artificial intelligence in inflammatory bowel disease, World J. Gastroenterol. 26 (2020), 6923–6928. https://doi.org/10.3748/wjg.v26.i44.6923

Q.A. Hathaway, S.M. Roth, M.V. Pinti, D.C. Sprando, A. Kunovac, A.J. Durr, et al., Machine-learning to stratify diabetic patients using novel cardiac biomarkers and integrative genomics, Cardiovasc. Diabetol. 18 (2019), 78. https://doi.org/10.1186/s12933-019-0879-0

V. Fortino, L. Wisgrill, P. Werner, S. Suomela, N. Linder, E. Jalonen, et al., Machine-learning–driven biomarker discovery for the discrimination between allergic and irritant contact dermatitis, Proc. Natl. Acad. Sci. U S A. 117 (2020), 33474–33485. https://doi.org/10.1073/pnas.2009192117

H. Johansson, M. Lindstedt, A.-S. Albrekt, C.A.K. Borrebaeck, A genomic biomarker signature can predict skin sensitizers using a cell-based in vitro alternative to animal tests, BMC Genomics. 12 (2011), 399. https://doi.org/10.1186/1471-2164-12-399

L.J. Marcos-Zambrano, K. Karaduzovic-Hadziabdic, T.L. Turukalo, P. Przymus, V. Trajkovik, O. Aasmets, et al., Applications of machine learning in human microbiome studies: a review on feature selection, biomarker identification, disease prediction and treatment, Front. Microbiol. 12 (2021), 634511. https://pubmed.ncbi.nlm.nih.gov/33737920/

H.S.R. Rajula, G. Verlato, M. Manchia, N. Antonucci, V. Fanos, Comparison of conventional statistical methods with machine learning in medicine: diagnosis, drug development, and treatment, Medicina. 56 (2020), 455. https://doi.org/10.3390/medicina56090455

M.G.H. Sanders, T. Nijsten, J. Verlouw, R. Kraaij, L.M. Pardo, Composition of cutaneous bacterial microbiome in seborrheic dermatitis patients: a cross-sectional study, PLoS ONE. 16 (2021), e0251136. https://doi.org/10.1371/journal.pone.0251136

E.H. Simpson, Measurement of diversity, Nature. 163 (1949), 688. https://doi.org/10.1038/163688a0

N.V. Chawla , K.W. Bowyer, L.O. Hall, W.P. Kegelmeyer, SMOTE: synthetic minority over-sampling technique, J. Artifi. Intellig. Res. 16 (2002), 321–357. https://doi.org/10.1613/jair.953

H.E.C. van der Wall, R.J. Doll, G.J.P. van Westen, I. Koopmans, R.G. Zuiker, J. Burggraaf, et al., The use of machine learning improves the assessment of drug-induced driving behaviour, Accid. Anal. Prevention. 148 (2020), 105822. https://doi.org/10.1016/j.aap.2020.105822

S. Lundberg, S.-I. Lee, A unified approach to interpreting model predictions, arXiv preprint, arXiv:1705.07874, 2017. https://doi.org/10.48550/arXiv.1705.07874

Z. Xu, Z. Wang, C. Yuan, X. Liu, F. Yang, T. Wang, et al., Dandruff is associated with the conjoined interactions between host and microorganisms, Sci. Rep. 6 (2016), 24877. https://doi.org/10.1038/srep24877

L.B. Nørreslet, T. Agner, M.-L. Clausen, The skin microbiome in inflammatory skin diseases, Curr. Derm. Rep. 9 (2020), 141–151. https://doi.org/10.1007/s13671-020-00297-z

M. Picardo, M. Ottaviani, Skin microbiome and skin disease: the example of rosacea, J. Clin. Gastroenterol. 48 (2014), S85–S86. https://doi.org/10.1097/mcg.0000000000000241

A.P. Carrieri, N. Haiminen, S. Maudsley-Barton, L.-J. Gardiner, B. Murphy, A.E. Mayes, et al., Explainable AI reveals changes in skin microbiome composition linked to phenotypic differences, Sci. Rep. 11 (2021), 4565. https://pubmed.ncbi.nlm.nih.gov/33633172/

S.K. Shukla, N.S. Murali, M.H. Brilliant, Personalized medicine going precise: from genomics to microbiomics, Trends Mol. Med. 21 (2015), 461–462. https://doi.org/10.1016/j.molmed.2015.06.002

J. Namkung, Machine learning methods for microbiome studies, J. Microbiol. 58 (2020), 206–216. https://doi.org/10.1007/s12275-020-0066-8

D. Knights, E.K. Costello, R. Knight, Supervised classification of human microbiota, FEMS Microbio. Reviews. 35 (2011), 343–359. https://doi.org/10.1111/j.1574-6976.2010.00251.x

Y.-H. Zhou, P. Gallins, A review and tutorial of machine learning methods for microbiome host trait prediction, Front. Genet. 10 (2019), 579. https://doi.org/10.3389/fgene.2019.00579

L. Moitinho-Silva, G. Steinert, S. Nielsen, C.C.P. Hardoim, Y.-C. Wu, G.P. McCormack, et al., Predicting the HMA-LMA status in marine sponges by machine learning, Front. Microbiol. 8 (2017), 752. https://doi.org/10.3389/fmicb.2017.00752

H.R. Johnson, D.D. Trinidad, S. Guzman, Z. Khan, J.V. Parziale, J.M. DeBruyn, et al., A machine learning approach for using the postmortem skin microbiome to estimate the postmortem interval, PLoS ONE. 11 (2016), e0167370. https://doi.org/10.1371/journal.pone.0167370

F. Tamer, M.E. Eren Yuksel, E. Sarifakioglu, Y. Karabag, Staphylococcus aureus is the most common bacterial agent of the skin flora of patients with seborrheic dermatitis, Dermatol. Pract. Concept. 8 (2018), 80–84. https://doi.org/10.5826/dpc.0802a04

L. Vicent, M. Martínez-Sellés, Electrocardiogeriatrics: ECG in advanced age, J. Electrocardio. 50 (2017), 698–700. https://doi.org/10.1016/j.jelectrocard.2017.06.003

A. Tanaka, O. Cho, C. Saito, M. Saito, R. Tsuboi, T. Sugita, Comprehensive pyrosequencing analysis of the bacterial microbiota of the skin of patients with seborrheic dermatitis, Microbiol. Immunol. 60 (2016), 521–526. https://doi.org/10.1111/1348-0421.12398

R.J. Hay, Malassezia, dandruff and seborrhoeic dermatitis: an overview, Br. J. Dermatol. 165 (2011), 2–8. https://doi.org/10.1111/j.1365-2133.2011.10570.x

A.K. Gupta, R. Batra, R. Bluhm, T. Boekhout, T.L. Dawson, Skin diseases associated with Malassezia species, J. Am. Acad. Dermatol. 51 (2004), 785–798. https://pubmed.ncbi.nlm.nih.gov/15523360/

J.L. Metcalf, Z.Z. Xu, S. Weiss, S. Lax, W.V. Treuren, E.R. Hyde, et al., Microbial community assembly and metabolic function during mammalian corpse decomposition, Science. 351 (2016), 158–162. https://pubmed.ncbi.nlm.nih.gov/26657285/

S. Seite, T. Bieber, Barrier function and microbiotic dysbiosis in atopic dermatitis, Clin. Cosmet. Investig. Dermatol. 8 (2015), 479–483. https://doi.org/10.2147/ccid.s91521



How to Cite

van der Wall H, Doll R, van Westen G, Niemeyer-van der Kolk T, Feiss G, Pinckaers H, van Doorn M, Nijsten T, Sanders M, Cohen A, Burggraaf J, Rissmann R, Pardo L. Discriminative Machine Learning Analysis for Skin Microbiome: Observing Biomarkers in Patients with Seborrheic Dermatitis. JAIMS [Internet]. 2022 Sep. 6 [cited 2023 Apr. 1];3(1-2):1-7. Available from: http://oapublishing-jaims.com/jaims/article/view/80