Machine learning to predict microbial community functions: An analysis of dissolved organic carbon from litter decomposition

Abstract

Microbial communities are ubiquitous and often influence macroscopic properties of the ecosystems they inhabit. However, deciphering the functional relationship between specific microbes and ecosystem properties is an ongoing challenge owing to the complexity of the communities. This challenge can be addressed, in part, by integrating the advances in DNA sequencing technology with computational approaches like machine learning. Although machine learning techniques have been applied to microbiome data, use of these techniques remains rare, and user-friendly platforms to implement such techniques are not widely available. We developed a tool that implements neural network and random forest models to perform regression and feature selection tasks on microbiome data. In this study, we applied the tool to analyze soil microbiome (16S rRNA gene profiles) and dissolved organic carbon (DOC) data from a 44-day plant litter decomposition experiment. The microbiome data includes 1709 total June 4, 2019 1/23 bacterial operational taxonomic units (OTU) from 300+ microcosms. Regression analysis of predicted and actual DOC for a held-out test set of 51 samples yield Pearson's correlation coefficients of .636 and .676 for neural network and random forest approaches, respectively. Important taxa identified by the machine learning techniques are compared to results from a standard tool (indicator species analysis) widely used by microbial ecologists. Of 1709 bacterial taxa, indicator species analysis identified 285 taxa as significant determinants of DOC concentration. Of the top 285 ranked features determined by machine learning methods, a subset of 86 taxa are common to all feature selection techniques. Using this subset of features, prediction results for random permutations of the data set are at least equally accurate compared to predictions determined using the entire feature set. Our results suggest that integration of multiple methods can aid identification of a robust subset of taxa within complex communities that may drive specific functional outcomes of interest. Introduction 1 Microbial communities mediate essential functions in diverse ecosystems. While the 2 microbiome controls many interesting macroscopic properties, elucidating the 3 relationship between specific microbes and ecosystem functions remains a complex 4 problem in ecology. Recent advances in DNA sequencing technology make it easy to 5 acquire metagenomic data representing the taxonomic profile of bacteria and fungi in 6 microbial communities. This opens the door to deciphering which components of the 7 microbiome can drive changes in macroscopic properties. However, analysis of 8 metagenomic microbial data poses several difficulties. The data are typically high 9 dimensional (many taxa) with a small number of samples collected in each study. 10 Additionally, sequencing results are noisy and yield sparse data sets [1]. 11 Machine learning techniques provide a means to analyze high-dimensional data [2, 3] 12 and could be used to elucidate relationships be...

Mechanistic insights into the attenuation of intestinal inflammation and modulation of the gut microbiome by krill oil using in vitro and in vivo models

“…Machine learning algorithms, such as random forests (RF), are excellent tools for classifying microbiome features into various classes or categories, and which also allow us to dissect the relationships between microbial features and environmental attributes. RF is less sensitive to the sample size of the training data set and more accurate for prediction performance [27]. The 20 most important genera selected by RF provided an accuracy classification between the KO and SO groups (Fig.…”

Microbial community composition controls carbon flux across litter types in early phase of litter decomposition

“…The Random Forest Indicator species analysis Neural Network (RFINN) platform (Thompson et al ., 2019) was used to identify subsets of genera that were found to be important features for prediction of either CO 2 or DOC. The data were pre‐processed with genera abundances computed as the average abundance of all OTUs within each genus.…”

Multi-omics study of keystone species in the cystic fibrosis lung microbiome

“…Samples were clustered using the proximity matrix from an unsupervised random forest in the R package randomForest using 5000 trees and Ward method for hierarchical clustering. Random forest is a robust non-parametric, statistical learning method for the identification of clusters and variables of importance in complex multi-omics data [54,55]. The effect of different antibiotic regimes (clindamycin, no antibiotics and the combination of colistin, piperacillin-tazobactam, ceftazidime-avibactam, and vancomycin) on was tested using a supervised permutational random Forest (5000 trees 1000 permutations) where the three antibiotic regimes were used as supervising variable in the package rfPermute.…”