Skip to contents

Feature summary

feature_summary.Rmd

Create Metaboprep object 

library(metaboprep)

# import data
data     <- read.csv(system.file("extdata", "dummy_data.csv",     package = "metaboprep"), header=T, row.names = 1) |> as.matrix()
samples  <- read.csv(system.file("extdata", "dummy_samples.csv",  package = "metaboprep"), header=T, row.names = 1)
features <- read.csv(system.file("extdata", "dummy_features.csv", package = "metaboprep"), header=T, row.names = 1)

# create object
mydata <- Metaboprep(data = data, samples = samples, features = features)

Summary of Metaboprep object 

summary(mydata)
#> Metaboprep Object Summary
#> --------------------------
#> Samples      : 100
#> Features     : 20
#> Data Layers  : 1
#> Layer Names  : input
#> 
#> Sample Summary Layers : none
#> Feature Summary Layers: none
#> 
#> Sample Annotation (metadata):
#>   Columns: 5
#>   Names  : sample_id, age, sex, pos, neg
#> 
#> Feature Annotation (metadata):
#>   Columns: 5
#>   Names  : feature_id, platform, pathway, derived_feature, xenobiotic_feature
#> 
#> Exclusion Codes Summary:
#> 
#>   Sample Exclusions:
#> Exclusion | Count
#> -----------------
#> user_excluded                     | 0
#> extreme_sample_missingness        | 0
#> user_defined_sample_missingness   | 0
#> user_defined_sample_totalpeakarea | 0
#> user_defined_sample_pca_outlier   | 0
#> 
#>   Feature Exclusions:
#> Exclusion | Count
#> -----------------
#> user_excluded                    | 0
#> extreme_feature_missingness      | 0
#> user_defined_feature_missingness | 0

Run feature summary 

# note that for illustrative purposes we are using a log outlier unit distance of 1.0 here, in practice we tend to favor a value of 5.0.
feature_sum1 <- feature_summary(metaboprep     = mydata, 
                                source_layer    = "input", 
                                outlier_udist   = 1.0,
                                tree_cut_height = 0.5,
                                output          = "data.frame")

Table of feature summary

feature_id missingness outlier_count n mean sd median min max range skew kurtosis se missing var disp_index coef_variance W log10_W k independent_features
metab_id_1 0 5 100 0.511 0.293 0.530 0.000 0.993 0.992 -0.123 -1.231 0.029 0 0.086 0.168 0.574 0.949 0.744 1 TRUE
metab_id_2 0 0 100 0.521 0.310 0.547 0.018 0.993 0.975 -0.150 -1.404 0.031 0 0.096 0.184 0.594 0.924 0.834 2 TRUE
metab_id_3 0 10 100 0.488 0.283 0.504 0.001 0.995 0.994 -0.036 -1.109 0.028 0 0.080 0.165 0.580 0.963 0.749 3 TRUE
metab_id_4 0 5 100 0.464 0.286 0.466 0.004 0.992 0.988 0.092 -1.199 0.029 0 0.082 0.177 0.617 0.954 0.833 4 TRUE
metab_id_5 0 11 100 0.521 0.293 0.547 0.004 0.976 0.972 -0.219 -1.161 0.029 0 0.086 0.164 0.561 0.945 0.782 5 TRUE
metab_id_6 0 7 100 0.490 0.259 0.473 0.007 0.993 0.986 0.007 -1.006 0.026 0 0.067 0.137 0.528 0.973 0.803 6 TRUE
metab_id_7 0 7 100 0.479 0.277 0.441 0.029 0.992 0.963 0.135 -1.211 0.028 0 0.077 0.160 0.579 0.953 0.899 7 TRUE
metab_id_8 0 0 100 0.476 0.312 0.491 0.001 0.999 0.998 0.059 -1.350 0.031 0 0.097 0.205 0.656 0.936 0.796 8 TRUE
metab_id_9 0 10 100 0.468 0.260 0.489 0.005 0.975 0.971 0.000 -1.090 0.026 0 0.068 0.144 0.556 0.968 0.800 9 TRUE
metab_id_10 0 0 100 0.524 0.290 0.532 0.019 0.993 0.974 -0.158 -1.252 0.029 0 0.084 0.161 0.554 0.945 0.841 10 TRUE

Feature summary attributes

In addition to the summary data, the hierarchical cluster dendrogram is appended to the returned data.frame as and attribute. This can be accessed with the attribute name: [source_layer]_tree, in this case we summarised the input data, therefore the attribute name is input_tree.

suppressPackageStartupMessages(library(dendextend))

# extract tree from attributes
tree <- attr(feature_sum1, 'input_tree')
dend <- stats::as.dendrogram(tree)

# color the independent features blue
metab_color       <- feature_sum1[, c("feature_id", "independent_features")]
metab_color       <- metab_color[match(labels(dend), metab_color$feature_id), ]
metab_color$color <- ifelse(metab_color$independent_features==TRUE, "#477EB8", "grey")

# format dendrogram for ploting
dend <- dend |>
  dendextend::set("labels_cex", 0.75) |>
  dendextend::set("labels_col", metab_color$color) |>
  dendextend::set("branches_lwd", 1) |>
  dendextend::set("branches_k_color",  value = metab_color$color)

## plot the dendrogram
dend |> plot(main = "Feature clustering dendrogram")
abline(h = 0.5, col = "#E41A1C", lwd = 1.5)

Decision tree showing feature importance in dataset

Run feature summary on subset

Using the sample_ids and feature_ids arguments you can run the summary for a subset of the data. Note: all rows will be return, however summary data will only be returned for the specified ids.

## define a vector of sample IDs
sids <- mydata@samples[mydata@samples$sex == "female", "sample_id"] 

## define a vector of feature IDs
fids <- mydata@features[, "feature_id"] |> sample(10)

# note that for illustrative purposes we are using a log outlier unit distance of 1.0 here, in practice we tend to favor a value of 5.0.
feature_sum_subset <- feature_summary(metaboprep      = mydata, 
                                      source_layer    = "input", 
                                      outlier_udist   = 1.0,
                                      tree_cut_height = 0.5,
                                      sample_ids      = sids,
                                      feature_ids     = fids,
                                      output          = "data.frame")

Table of feature summary for subset

feature_id missingness outlier_count n mean sd median min max range skew kurtosis se missing var disp_index coef_variance W log10_W k independent_features
metab_id_4 0 5 55 0.498 0.273 0.478 0.009 0.992 0.983 0.094 -1.159 0.037 0 0.075 0.150 0.549 0.960 0.836 7 TRUE
metab_id_5 0 9 55 0.580 0.296 0.632 0.013 0.976 0.963 -0.436 -1.035 0.040 0 0.087 0.151 0.510 0.929 0.734 3 TRUE
metab_id_6 0 6 55 0.520 0.284 0.560 0.007 0.993 0.986 -0.273 -1.103 0.038 0 0.080 0.155 0.546 0.952 0.757 4 TRUE
metab_id_9 0 6 55 0.461 0.270 0.489 0.005 0.975 0.971 0.012 -1.151 0.036 0 0.073 0.158 0.585 0.963 0.798 9 TRUE
metab_id_12 0 0 55 0.518 0.296 0.498 0.006 0.989 0.983 -0.125 -1.233 0.040 0 0.088 0.170 0.573 0.946 0.801 2 TRUE
metab_id_13 0 4 55 0.544 0.313 0.595 0.005 0.990 0.984 -0.265 -1.325 0.042 0 0.098 0.181 0.576 0.927 0.752 1 TRUE
metab_id_15 0 7 55 0.495 0.280 0.522 0.032 0.991 0.959 0.054 -1.139 0.038 0 0.078 0.158 0.566 0.961 0.874 5 TRUE
metab_id_16 0 3 55 0.463 0.315 0.447 0.005 0.998 0.992 0.276 -1.384 0.042 0 0.099 0.214 0.679 0.909 0.868 6 TRUE
metab_id_17 0 3 55 0.449 0.287 0.437 0.025 0.972 0.947 0.185 -1.280 0.039 0 0.083 0.184 0.639 0.938 0.902 8 TRUE
metab_id_18 0 1 55 0.536 0.316 0.599 0.028 0.999 0.971 -0.073 -1.480 0.043 0 0.100 0.186 0.590 0.918 0.867 10 TRUE

Run sample & feature summaries together 

# note that for illustrative purposes we are using a log outlier unit distance of 1.0 here, in practice we tend to favor a value of 5.0.
sam_n_feat_sum <- summarise(metaboprep      = mydata, 
                            source_layer    = "input", 
                            outlier_udist   = 1.0,
                            tree_cut_height = 0.5,
                            sample_ids      = sids, 
                            feature_ids     = fids,
                            output          = "data.frame")

Table of feature summary for subset

feature_id missingness outlier_count n mean sd median min max range skew kurtosis se missing var disp_index coef_variance W log10_W k independent_features
metab_id_4 0 5 55 0.498 0.273 0.478 0.009 0.992 0.983 0.094 -1.159 0.037 0 0.075 0.150 0.549 0.960 0.836 7 TRUE
metab_id_5 0 9 55 0.580 0.296 0.632 0.013 0.976 0.963 -0.436 -1.035 0.040 0 0.087 0.151 0.510 0.929 0.734 3 TRUE
metab_id_6 0 6 55 0.520 0.284 0.560 0.007 0.993 0.986 -0.273 -1.103 0.038 0 0.080 0.155 0.546 0.952 0.757 4 TRUE
metab_id_9 0 6 55 0.461 0.270 0.489 0.005 0.975 0.971 0.012 -1.151 0.036 0 0.073 0.158 0.585 0.963 0.798 9 TRUE
metab_id_12 0 0 55 0.518 0.296 0.498 0.006 0.989 0.983 -0.125 -1.233 0.040 0 0.088 0.170 0.573 0.946 0.801 2 TRUE
metab_id_13 0 4 55 0.544 0.313 0.595 0.005 0.990 0.984 -0.265 -1.325 0.042 0 0.098 0.181 0.576 0.927 0.752 1 TRUE
metab_id_15 0 7 55 0.495 0.280 0.522 0.032 0.991 0.959 0.054 -1.139 0.038 0 0.078 0.158 0.566 0.961 0.874 5 TRUE
metab_id_16 0 3 55 0.463 0.315 0.447 0.005 0.998 0.992 0.276 -1.384 0.042 0 0.099 0.214 0.679 0.909 0.868 6 TRUE
metab_id_17 0 3 55 0.449 0.287 0.437 0.025 0.972 0.947 0.185 -1.280 0.039 0 0.083 0.184 0.639 0.938 0.902 8 TRUE
metab_id_18 0 1 55 0.536 0.316 0.599 0.028 0.999 0.971 -0.073 -1.480 0.043 0 0.100 0.186 0.590 0.918 0.867 10 TRUE