10.6 Microbiome Single-Omics Quick-Start Example

This section demonstrates a complete and commonly used analytical workflow for microbiome single-omics data using example datasets.

Example data download: Github link

10.6.1 Importing Microbiome Data

library(EasyMultiProfiler)

meta_data <- read.table('coldata.txt',header = T,row.names = 1)
data <- read.table('tax.txt',header = T,sep = '\t')

MAE <- EMP_easy_import(data = data,coldata = meta_data,type = 'tax')

10.6.2 Exploring Microbiome Data

View Current Microbiome Assay

MAE |>
  EMP_assay_extract() # View expression matrix

MAE |>
  EMP_coldata_extract() # View phenotype data

MAE |>
  EMP_rowdata_extract() # View taxonomic annotations

View Phylum-Level Data

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Phylum')

View Class-Level Data

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Class')

10.6.3 Rarefaction (Optional)

Rarefy Using the Smallest Read Count Among Samples

MAE |>
  EMP_assay_extract() |>
  EMP_rrarefy()

Rarefy with a Custom Minimum Read Count

MAE |>
  EMP_assay_extract() |>
  EMP_rrarefy(raresize = 5000)

10.6.4 Data Normalization

Convert to Relative Abundance at Genus Level

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_decostand(method = 'relative')

Apply CLR Transformation at Genus Level

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_decostand(method = 'clr')

Apply Log2 Transformation at Genus Level

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_decostand(method = 'log2+1')

10.6.5 Batch Effect Correction (Optional)

Correct for Batch Effects by Region at Genus Level

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_adjust_abundance(.factor_unwanted = 'Region',
                       .factor_of_interest = 'Group',
                       method = 'combat_seq')

10.6.6 Core Microbiome Identification

Identify Core Genera with Minimum Abundance 0.001 and Prevalence >70% in at Least One Group

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_identify_assay(estimate_group = 'Group',method = 'default',
                     min=0.001,min_ratio = 0.7)

10.6.7 Alpha Diversity Analysis

Calculate Alpha Diversity for Core Genera and Visualize with Boxplot

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_identify_assay(estimate_group = 'Group',method = 'default',
                     min=0.001,min_ratio = 0.7) |>
  EMP_alpha_analysis() |>
  EMP_boxplot(estimate_group = 'Group')

10.6.8 Beta Diversity Analysis

Calculate Beta Diversity for Core Genera and Generate Ordination Plot

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_identify_assay(estimate_group = 'Group',method = 'default',
                     min=0.001,min_ratio = 0.7) |>
  EMP_dimension_analysis(method = 'pcoa',distance = 'bray') |>
  EMP_scatterplot(estimate_group = 'Group',show='p12html')

10.6.9 Differential Abundance Analysis

Perform Wilcoxon Test and Filter Significant Genera

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_diff_analysis(method = 'wilcox.test',estimate_group = 'Group') |>
  EMP_filter(feature_condition = pvalue < 0.05)

Perform DESeq2 Analysis and Filter Significant Genera

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_diff_analysis(method = 'DESeq2',.formula = ~Group) |>
  EMP_filter(feature_condition = pvalue < 0.05,keep_result = TRUE)

10.6.10 Machine Learning for Key Taxa

The EMP package includes built-in methods for feature selection: Boruta, Random Forest, XGBoost, and Lasso. For detailed usage, run help(EMP_marker_analysis).

Identify Important Genera Using Boruta and Visualize with Heatmap

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_identify_assay(estimate_group = 'Group',method = 'default',
                     min=0.001,min_ratio = 0.7) |>
  EMP_marker_analysis(method = 'boruta',estimate_group = 'Group') |>
  EMP_filter(feature_condition = Boruta_decision!= 'Rejected') |>
  EMP_heatmap_plot(palette='Spectral',legend_bar='auto',
                   clust_row=TRUE,clust_col=TRUE)

Select Height-Associated Genera Using Lasso and Plot Group-Level Heatmap

MAE |>
  EMP_assay_extract() |> 
  EMP_collapse(collapse_by = 'row',estimate_group = 'Species') |>
  EMP_identify_assay(estimate_group = 'Group',method = 'default',
                     min=0.001,min_ratio = 0.7) |>
  EMP_marker_analysis(method = 'lasso',estimate_group = 'Height') |>
  EMP_filter(feature_condition = lasso_coe > 0) |>
  EMP_collapse(method = 'mean',estimate_group = 'Group',
               collapse_by = 'col') |>
  EMP_heatmap_plot(palette='Spectral',legend_bar='auto')

10.6.11 Correlation with Phenotypes

Correlation Heatmap Between Phyla and Phenotypic Variables

phylum_data <- MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Phylum')

meta_data <- MAE |>
  EMP_coldata_extract(action = 'add')

(phylum_data + meta_data) |>
  EMP_cor_analysis() |>
  EMP_heatmap_plot()

10.6.12 Linear Regression with Phenotypes

Linear Fit Between Genus Blautia and BMI

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(collapse_by = 'row',estimate_group = 'Genus') |>
  EMP_fitline_plot(var_select=c('Blautia','BMI'))

10.6.13 Network Analysis

Network Plot Using Differentially Abundant Genera and Selected Phenotypes

MAE |>
  EMP_assay_extract() |>
  EMP_collapse(estimate_group = 'Genus',collapse_by = 'row') |>
  EMP_diff_analysis(method='wilcox.test', estimate_group = 'Group') |>
  EMP_filter(feature_condition = pvalue<0.05) |>
  EMP_network_analysis(coldata_to_assay = c('BMI','PHQ9','GAD7')) |>
  EMP_network_plot(node_info = 'Phylum',label.cex = 1,edge.labels = TRUE)

10.6 Microbiome example