Feature summary

library(omiprep)

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

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

summary(mydata)
#> Omiprep Object Summary
#> --------------------------
#> Samples      : 125
#> Features     : 253
#> Data Layers  : 1
#> Layer Names  : input
#> 
#> Sample Summary Layers : none
#> Feature Summary Layers: none
#> 
#> Sample Annotation (metadata):
#>   Columns: 10
#>   Names  : sample_id, parent_sample_id, client_identifier, sex, age, bmi, LC.MS.Polar, LC.MS.Neg, LC.MS.Pos.Early, LC.MS.Pos.Late
#> 
#> Feature Annotation (metadata):
#>   Columns: 14
#>   Names  : feature_id, pathway_sortorder, biochemical, super_pathway, sub_pathway, comp_id, platform, chemical_id, ri, mass, cas, pubchem, kegg, group_hmdb
#> 
#> 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
#> user_defined_feature_skewness    | 0

mydata = quality_control(mydata)
#> ── Starting Omics QC Process ───────────────────────────────────────────────────
#> ℹ Validating input parameters
#> ✔ Validating input parameters [16ms]
#> 
#> ℹ Sample & Feature Summary Statistics for raw data
#> ℹ Number of informative PCs (Scree acceleration factor): 2
#> ℹ Sample & Feature Summary Statistics for raw data✔ Sample & Feature Summary Statistics for raw data [2.4s]
#> 
#> ℹ Copying input data to new 'qc' data layer
#> ✔ Copying input data to new 'qc' data layer [22ms]
#> 
#> ℹ Assessing for extreme sample missingness >=80% - excluding 0 sample(s)
#> ✔ Assessing for extreme sample missingness >=80% - excluding 0 sample(s) [21ms]
#> 
#> ℹ Assessing for extreme feature missingness >=80% - excluding 0 feature(s)
#> ✔ Assessing for extreme feature missingness >=80% - excluding 5 feature(s) [21m…
#> 
#> ℹ Assessing for sample missingness at specified level of >=20% - excluding 0 sa…
#> ✔ Assessing for sample missingness at specified level of >=20% - excluding 1 sa…
#> 
#> ℹ Assessing for feature missingness at specified level of >=20% - excluding 0 f…
#> ✔ Assessing for feature missingness at specified level of >=20% - excluding 46 …
#> 
#> ℹ Calculating total sum abundance outliers at +/- 5 Sdev - excluding 0 sample(s)
#> ✔ Calculating total sum abundance outliers at +/- 5 Sdev - excluding 0 sample(s…
#> 
#> ℹ Running sample data PCA outlier analysis at +/- 5 Sdev
#> ✔ Running sample data PCA outlier analysis at +/- 5 Sdev [20ms]
#> 
#> ℹ Sample PCA outlier analysis - re-identify feature independence and PC outlier…
#> ℹ Number of informative PCs (Scree acceleration factor): 2
#> ℹ Sample PCA outlier analysis - re-identify feature independence and PC outlier…! The stated max PCs [max_num_pcs=10] to use in PCA outlier assessment is greater than the number of available informative PCs [2]
#> ℹ Sample PCA outlier analysis - re-identify feature independence and PC outlier…✔ Sample PCA outlier analysis - re-identify feature independence and PC outlier…
#> 
#> ℹ Creating final QC dataset...
#> ℹ Number of informative PCs (Scree acceleration factor): 2
#> ℹ Creating final QC dataset...
#> ℹ Creating final QC dataset...── Step timings ──
#> ℹ Creating final QC dataset...
#> ℹ Creating final QC dataset...
#>                         step seconds   pct
#>                   validation    0.00   0.0
#>                summarise_raw    2.36  38.8
#>                   copy_layer    0.00   0.0
#>   extreme_sample_missingness    0.00   0.0
#>  extreme_feature_missingness    0.00   0.0
#>           sample_missingness    0.00   0.0
#>          total_sum_abundance    0.01   0.2
#>                summarise_pca    1.74  28.6
#>              summarise_final    1.74  28.6
#>                        total    6.09 100.0
#> ✔ Creating final QC dataset... [1.8s]
#> 
#> ℹ 'Omics QC Process Completed
#> ✔ 'Omics QC Process Completed [17ms]

## The attributes include column names, row names, and class for the feature summary table
## as well as a hierarchical cluster dendrogram or `input_tree` and the parameter values for 
## outlier_udist and input_tree_cut_height passed to the function. 
names( attributes(feature_sum1) )
#> [1] "names"                 "row.names"             "class"                
#> [4] "input_tree"            "input_outlier_udist"   "input_tree_cut_height"

# NOTE:
# outlier_udist = number of IQRs from the median at which a value is flagged.
# 1.0 here is illustrative; in practice we favour 5.0, which is the default value
# for the quality_control() function.
sf_sum <- summarise(omiprep         = mydata, 
                    source_layer    = "input", 
                    outlier_udist   = 1.0,
                    tree_cut_height = 0.5,
                    sample_ids      = sids, ## It is also possible to run on a subset of samples and/or features
                    feature_ids     = fids,
                    output          = "data.frame", 
                    cores           = 1)
#> ℹ Number of informative PCs (Scree acceleration factor): 2

sf_sum$feature_summary |> 
  na.omit() |>
  knitr::kable( digits = 3, row.names = FALSE, align = "c") |>
  kableExtra::kable_styling(full_width = TRUE) |>
  kableExtra::scroll_box(width = "100%", height = "500px")