6533b827fe1ef96bd1285766

RESEARCH PRODUCT

Predicting pesticide biodegradation potential from microbial community composition: new tools for bioremediation

subject

description

Bioaugmentation is receiving increasing attention as a green technology to treat contaminatedareas by inoculating specific biodegrading microorganisms. However, our understanding of therole of microbial community composition and structure in the expression of contaminantdegradation potential is yet to improve. It could help making wise choice for microorganisms –community or specific strain – to be inoculated in contaminated soils with consideration to theirindigeneous microbiota.Here we tried to predict the microbial degradation of two herbicides, glyphosate andisoproturon by means of penalized regression and machine learning methods routinely used ingenomic selection. To this end, we conducted experimental modifications of these twoherbicides degrading communities by applying biocide treatments coupled with serial dilutions.We then applied three selected genomic selection methods (i.e. Ridge Regression, LASSO andRandom Forest) on these community variants to link their OTUs composition to their herbicidedegradation capacities.Resulting predictions power is compelling with more than 80% correlation between predictedand actual herbicide degradation capacities. Moreover, OTUs detected as having an impact onherbicide degradation were confronted to literature and validated. To go further and test therobustness of our methods, an experimental validation of the theoretical prediction was set up.Mixed resulting prediction quality calls for promising advances in the field of soil bioremediationwith the need of further improvement. Finally, futures applications in a bioremediationperspective are considered.