Get Started with glyexp • glyexp

In this context, you typically work with three types of data in glycomics and glycoproteomics experiments:

Expression data (assay) - the actual measurements of your biological molecules (glycans, glycopeptides, etc.)
Molecular annotations (row data) - the identifiers for your molecules (structures, sequences, etc.)
Experimental metadata (column data) - the context of your samples (time points, treatments, experimental conditions)

The SummarizedExperiment class in the SummarizedExperiment packag serves as a structured container that keeps all three data types organized and interconnected. glyexp provides two subclasses of SummarizedExperiment: GlycomicSE for glycomics data, and GlycoproteomicSE for glycoproteomics data.

For those familiar with SummarizedExperiment

GlycomicSE and GlycoproteomicSE are both subclasses of SummarizedExperiment with a few restrictions:

GlycomicSE requires a glycan_composition column as a glyrepr::glycan_composition() vector in the rowData.
GlycoproteomicSE requires glycan_composition, as well as protein (character) and protein_site (integer) columns.
There can be only one assay.

That’s all. All your familiar operations provided by SummarizedExperiment and tidySummarizedExperiment can also be used on GlycomicSE and GlycoproteomicSE.

Getting Started with glyexp

Let’s begin with a simple example to illustrate the basic concepts.

# `real_experiment` is a built-in `GlycoproteomicSE` object
exp <- real_experiment
exp
#> 
#> ── GlycoproteomicSE ────────────────────────────────────────────────────────────
#> ℹ Abundance assay: 12 samples, 4262 variables
#> ℹ Glycan type: N
#> ℹ Row data fields: peptide <chr>, peptide_site <int>, protein <chr>, protein_site <int>, gene <chr>, glycan_composition <comp>, glycan_structure <struct>
#> ℹ Column data fields: group <fct>
#> ℹ Metadata fields: exp_type <chr>, glycan_type <chr>, quant_method <chr>

The summary provides an overview of your entire experiment - variables, observations, and all the metadata.

The Expression Matrix

The expression matrix contains your numerical data - rows are variables (molecules), columns are observations (samples).

assay(exp)[1:5, 1:3]
#>                                              C1         C2           C3
#> P08185-176-Hex(5)HexNAc(4)NeuAc(2)           NA         NA     10655.62
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-1  414080036  609889761  78954431.49
#> P04196-344-Hex(5)HexNAc(4)            581723113  604842244 167889901.32
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-2 3299649335 2856490652 957651065.86
#> P10909-291-Hex(6)HexNAc(5)-1           30427048   34294394   6390129.81

Variable Information

The variable information table (or “raw data”, in the SummarizedExperiment terminology) contains detailed annotations for each molecule.

head(rowData(exp))
#> DataFrame with 6 rows and 7 columns
#>                                             peptide peptide_site     protein
#>                                         <character>    <integer> <character>
#> P08185-176-Hex(5)HexNAc(4)NeuAc(2)           NKTQGK            1      P08185
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-1 HSHNNNSSDLHPHK            5      P04196
#> P04196-344-Hex(5)HexNAc(4)           HSHNNNSSDLHPHK            5      P04196
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-2 HSHNNNSSDLHPHK            5      P04196
#> P10909-291-Hex(6)HexNAc(5)-1               HNSTGCLR            2      P10909
#> P04196-344-Hex(5)HexNAc(4)NeuAc(2)   HSHNNNSSDLHPHK            5      P04196
#>                                      protein_site        gene
#>                                         <integer> <character>
#> P08185-176-Hex(5)HexNAc(4)NeuAc(2)            176    SERPINA6
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-1          344         HRG
#> P04196-344-Hex(5)HexNAc(4)                    344         HRG
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-2          344         HRG
#> P10909-291-Hex(6)HexNAc(5)-1                  291         CLU
#> P04196-344-Hex(5)HexNAc(4)NeuAc(2)            344         HRG
#>                                          glycan_composition
#>                                       <glyrepr_composition>
#> P08185-176-Hex(5)HexNAc(4)NeuAc(2)   Hex(5)HexNAc(4)NeuAc..
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-1 Hex(5)HexNAc(4)NeuAc..
#> P04196-344-Hex(5)HexNAc(4)                  Hex(5)HexNAc(4)
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-2 Hex(5)HexNAc(4)NeuAc..
#> P10909-291-Hex(6)HexNAc(5)-1                Hex(6)HexNAc(5)
#> P04196-344-Hex(5)HexNAc(4)NeuAc(2)   Hex(5)HexNAc(4)NeuAc..
#>                                            glycan_structure
#>                                         <glyrepr_structure>
#> P08185-176-Hex(5)HexNAc(4)NeuAc(2)   NeuAc(??-?)Hex(??-?)..
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-1 NeuAc(??-?)Hex(??-?)..
#> P04196-344-Hex(5)HexNAc(4)           Hex(??-?)HexNAc(??-?..
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-2 NeuAc(??-?)Hex(??-?)..
#> P10909-291-Hex(6)HexNAc(5)-1         Hex(??-?)HexNAc(??-?..
#> P04196-344-Hex(5)HexNAc(4)NeuAc(2)   NeuAc(??-?)Hex(??-?)..

Sample Information

The sample information table (or “column data”) records the experimental conditions for each sample.

head(colData(exp))
#> DataFrame with 6 rows and 1 column
#>       group
#>    <factor>
#> C1        C
#> C2        C
#> C3        C
#> H1        H
#> H2        H
#> H3        H

Rownames and colnames

Row names in rowData() and colData() match the row and column names in the assay, repectively.

identical(rownames(rowData(exp)), rownames(assay(exp)))
#> [1] TRUE
identical(rownames(colData(exp)), colnames(assay(exp)))
#> [1] TRUE

As a shorthand, you can just use rownames() and colnames() on the container.

rownames(exp)[1:5]
#> [1] "P08185-176-Hex(5)HexNAc(4)NeuAc(2)"  
#> [2] "P04196-344-Hex(5)HexNAc(4)NeuAc(1)-1"
#> [3] "P04196-344-Hex(5)HexNAc(4)"          
#> [4] "P04196-344-Hex(5)HexNAc(4)NeuAc(1)-2"
#> [5] "P10909-291-Hex(6)HexNAc(5)-1"
colnames(exp)[1:5]
#> [1] "C1" "C2" "C3" "H1" "H2"

Data Manipulation with glyexp

glyexp provides dplyr-style functions for manipulating SummarizedExperiment objects.

For every dplyr function, glyexp provides two specialized versions:

_col() functions: work on your sample metadata
_row() functions: work on your variable annotations

Here’s an example of filtering for group “A” samples:

subset_exp <- filter_col(exp, group == "H")

Let’s check what happened to our sample info:

colData(subset_exp)
#> DataFrame with 3 rows and 1 column
#>       group
#>    <factor>
#> H1        H
#> H2        H
#> H3        H

Check the expression matrix:

assay(subset_exp)[1:5,]
#>                                                H1         H2           H3
#> P08185-176-Hex(5)HexNAc(4)NeuAc(2)   3.105412e+04         NA     457398.3
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-1           NA   11724908    3892287.5
#> P04196-344-Hex(5)HexNAc(4)           6.977076e+08  703566323  421872371.5
#> P04196-344-Hex(5)HexNAc(4)NeuAc(1)-2 2.600523e+09 3229968280 2653736044.6
#> P10909-291-Hex(6)HexNAc(5)-1         5.159133e+07   37479075   62332706.5

The expression matrix is automatically filtered to match!

This is filter_col(): it filters the sample information and automatically updates the expression matrix to match.

Variable filtering works the same way:

exp |> filter_col(group == "H")
#> 
#> ── GlycoproteomicSE ────────────────────────────────────────────────────────────
#> ℹ Abundance assay: 3 samples, 4262 variables
#> ℹ Glycan type: N
#> ℹ Row data fields: peptide <chr>, peptide_site <int>, protein <chr>, protein_site <int>, gene <chr>, glycan_composition <comp>, glycan_structure <struct>
#> ℹ Column data fields: group <fct>
#> ℹ Metadata fields: exp_type <chr>, glycan_type <chr>, quant_method <chr>

Notice how these functions support the pipe operator (|>)? That’s the dplyr DNA in action!

The pattern is straightforward: glyexp functions expect and return SummarizedExperiment objects, including GlycomicSE and GlycoproteomicSE, and they preserve the row and column identifiers during operations.

To learn more about dplyr-style operations, check out this vignette. To learn basic SummarizedExperiment operations, check out this vignette.

Background and Design Principles

glyexp builds on insights from related data containers:

SummarizedExperiment The foundational omics data container from Bioconductor. Well-established for RNA-seq analysis.

tidySummarizedExperiment An attempt to bring tidy principles to SummarizedExperiment from the tidySummarizedExperiment package. While the concept is sound, storing all components in a single tibble doesn’t align with the mental model of separated data types.

massdataset A related package for mass spectrometry data. It provides tidy operations, clean data separation, and data processing history tracking. We appreciate its approach to reproducibility.

While object-oriented programming has its merits, glyexp takes a functional programming approach. Your analysis code serves as the reproducibility record - clear, transparent, and familiar to R users.

Design Philosophy glyexp uses functional programming because it aligns with how R users work. The design emphasizes clear, chainable functions.

Thank you to all the developers who contributed to these foundational packages.

What’s Next?

Now you have the basic understanding of glyexp. Next, you can learn how to use it in your analysis.