The aim of this package is to supply easy extraction and manipulation of SBML information by insertion in tabular data structures. Because of descriptive nature of SBML documents, the dataframe format is particularly suitable for easily access data and be able to perform subsequent analysis. Specially, this type of conversion enables easy data interrogation by means of tidyverse verbs in order to facilitate, for instance, usage of biomaRt and igraph-like packages. The involvement in the Bioconductor project establishes a direct and consistent connection with bioinformatics community while providing cooperation of tools useful also within the frame of systems biology and, in general, for the analysis of biological data.

In order to illustrate the package functioning, we used as examples an SBML file (Hucka et al. 2019) extracted from Reactome (Milacic et al. 2023), an open-source, open-access and peer-reviewed biological pathway database. Namely, the pathway is the “Aryl hydrocarbon receptor signalling” (R-HSA-8937144 (Jassal 2016)).

After providing installation instructions, the first section describes the dataframes structure, in the subsequent two sections are described the tidysbml steps to follow for pursuing the SBML conversion, while in the last one are shown some examples to integrate tidysbml dataframes with other Bioconductor packages (i.e. biomaRt and RCy3). In the following, it is useful to distinguish the SBML tags names using italic and the R commands with teletype fonts, respectively. Also, the terms ‘tag’ and ‘component’ are used interchangeably.

Installation

To install tidysbml from Bioconductor, run

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("tidysbml")

General dataframes structure

The SBML components of main interest for this package are listOfCompartments, listOfSpecies and listOfReactions. Due to the underlying SBML format, these dataframes may generally consist of the following three sets of columns: (i) Attributes, involving tags such as id, metaid, name, etc., (ii) Notes, consisting only of notes tag and (iii) Annotation, whose qualifiers are tags like bqmodel:is, bqbiol:is, bqbiol:hasPart, etc. Each tag is exported in one separate column. Columns for tags in the Attributes set are named as the tag name, notes column is named ‘notes’, while Annotation columns are prefixed by ‘annotation_’ followed by tag name after colon symbol (‘:’), for instance column with bqbiol:is tag content is labelled ‘annotation_is’. If one entity possesses multiple tags with the same name, the repeated column name is accompanied by a number (from the second copy it starts from ’_1’). Whether more values are contained in one tag (e.g. as happens for Annotation tags such as bqbiol:hasPart, bqbiol:isDescribedBy, or also Notes column) they are separated by delimiters like " " (i.e. single space character) for Annotation values and “|” (i.e. pipe character) for Notes. Also, based on the selected component, the respective dataframe may contain other columns, depending on the xml structure of the underlying component’s class. See for instance the df_species_in_reactions dataframe described in the following.

Converting SBML into a R list

The first step to convert an SBML file into R dataframes is to convert the SBML document into an R list object, by means of the sbml_as_list() function. In fact, all the other functions in this package require a list as input. The sole exception is given by as_dfs() which incorporates this conversion function and therefore may also receive directly an SBML file as first argument (and not exclusively an SBML converted into a list object).

The sbml_as_list() function exploits functions for reading and converting xml files from the xml2 R package (Wickham, Hester, and Ooms 2023) and outputs an appropriate type of list. In the following a such list is referred as SBML-converted list. The first argument reads the file path where the SBML file is located, the second one sets the information about which is the SBML component the user wants to look at (i.e. among the listOfSpecies, listOfCompartments and listOfReactions), if any (the default option gets ‘all’ components). Examples for both options are given below.

After running

library(tidysbml)

example of default option is

filepath <- system.file("extdata", "R-HSA-8937144.sbml", package = "tidysbml")
sbml_list <- sbml_as_list(filepath)

that returns a full SBML model, starting from the sbml tag, converted into a list of lists nested accordingly to the xml nesting rules.

Instead, an example of SBML-list conversion for only the list of species is given by

list_species <- sbml_as_list(filepath, component = "species")

which yields an SBML-converted list of lists starting from the listOfSpecies tag, that is contained inside the sbml and model tags. This last output is required in case the user is interested in the extraction of only one dataframe, e.g., using the as_df() function, as described in the following section.

Converting SBML into dataframes

The main function of this package is as_dfs(), which is able to provide the SBML information about Compartments, Species and Reactions in a tabular format. It returns a list of at most 4 dataframes, depending on the components reported inside the SBML selected.

The dataframe for listOfCompartments (listOfSpecies) component, named df_compartments (df_species), has one row for each Compartment (Species) and one column for each Attributes, Notes and Annotation value. Similarly, the first dataframe about Reactions (i.e. df_reactions) contains one row for each Reaction with their Attributes, Notes and Annotation values as columns, while the second one (i.e. df_species_in_reactions) has one row for each Species involved in each Reaction, here with the addition of two more columns: the reaction_id column, with information about the corresponding Reaction identificator reported in the SBML document, and the container_list column, with the name of the listOf element containing that Species (i.e. listOfReactants, listOfProducts, listOfModifiers). It is possible to use as first argument the SBML file path

list_of_dfs <- as_dfs(filepath, type = "file")
#> Empty notes' column for 'compartment' elements

or directly the SBML-converted list after using sbml_as_list() as described above

list_of_dfs <- as_dfs(sbml_list, type = "list")
#> Empty notes' column for 'compartment' elements

both returning the same output, that is the list with all the dataframes available from extraction. After the list has been extracted once, this second way is preferable, in order to avoid repeated sbml-list conversions.

Another function, namely as_df(), enables the conversion of only one dataframe at a time, depending on the SBML component of interest. Here a SBML-converted list starting from listOfCompartments/listOfSpecies/listOfReactions component is a mandatory input. For instance, converting first the SBML file into a list focusing at the listOfSpecies component

list_species <- sbml_as_list(filepath, component = "species")
df_species <- as_df(list_species)

returns one dataframe containing all information about species. Just for listOfReactions component is possible to obtain two dataframes. Here dfs_about_reactions is a list of 2 dataframes obtained by

list_react <- sbml_as_list(filepath, component = "reactions")
dfs_about_reactions <- as_df(list_react)

whose first component, containing information about reactions, returns 15 columns for the 5 reactions of our example

dfs_about_reactions[[1]]

While df_species_in_reactions, with information about the 14 species involved in the 5 reactions described above, is obtained by taking the second component

dfs_about_reactions[[2]]

Each function described in this section performs a control on the input format correctness. In particular, it returns errors if the input object is an empty list or not a list object, and also if its format is not suitable for extraction (i.e. SBML tags are not properly named or nested). In particular, the SBML file is accepted by as_df() if it contains only one type of tag within the first level of ‘listOf’ components. For instance, if the SBML is restricted to listOfSpecies/listOfCompartments/listOfReactions tag, the only type of tag within the list should be species/compartment/reaction. One more condition, given only in the as_dfs() function, is that the first two tags in the xml hierarchy should be sbml and model, where the former contains the latter. If any one of these conditions does not hold, the respective functions are not executed.

Integration with Bioconductor packages

This section provides R code to incorporate tidysbml dataframes with other Bioconductor packages. Here are shown examples of integration for RCy3 (Gustavsen et al. 2019) and biomaRt (Durinck et al. 2005) packages.

RCy3 package permits communication between R and Cytoscape softwares. After launching Cytoscape, it is possible to import graph in form of edgelist (i.e. dataframe with source and target columns) by simple (or heavier) data manipulation through dataframes as

library(dplyr)
edgelist <- df_species_in_reactions %>% select("reaction_id", "species") %>% `colnames<-`(c("source", "target"))
RCy3::createNetworkFromDataFrames(edges = edgelist) # while running Cytoscape

BiomaRt, instead, is an annotation package providing access to external public databases. One possible usage, for instance, is to visualize information about Uniprot ids reported in SBML for Species, here considering only those composed by multiple entities (i.e. multiple ids). First, extract URIs data about species from Annotation column with bqbiol:hasPart content

vec_uri <- na.omit( unlist(
  lapply(X = list_of_dfs[[2]]$annotation_hasPart, FUN = function(x){
    unlist(strsplit(x, "||", fixed = TRUE))
  })
))

filter only Uniprot URIs

vec_uniprot <- na.omit( unlist(
  lapply( X = vec_uri, FUN = function(x){
    if( all(unlist(gregexpr("uniprot", x)) > -1) ){
      x
    } else {
      NA
    }
  })
))

and extract Uniprot ids

vec_ids <- vapply(vec_uniprot, function(x){
  chr <- "/"
  first <- max(unlist(gregexpr(chr, x)))
  substr(x, first + 1, nchar(x))
}, FUN.VALUE = character(1))

Then, using biomaRt commands, user can set attributes information to look at

library(biomaRt)
mart <- useEnsembl(biomart = "ensembl", dataset = "hsapiens_gene_ensembl")
df_mart_uniprot <- getBM( attributes = c("uniprot_gn_id", "uniprot_gn_symbol",   "description"),
                          filters = "uniprot_gn_id",
                          values = vec_ids,
                          mart = mart)
df_mart_uniprot

Session info

sessionInfo()
#> R version 4.4.1 (2024-06-14)
#> Platform: x86_64-pc-linux-gnu
#> Running under: Ubuntu 24.04.1 LTS
#> 
#> Matrix products: default
#> BLAS:   /media/volume/teran2_disk/biocbuild/bbs-3.20-bioc/R/lib/libRblas.so 
#> LAPACK: /usr/lib/x86_64-linux-gnu/lapack/liblapack.so.3.12.0
#> 
#> locale:
#>  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
#>  [3] LC_TIME=en_GB              LC_COLLATE=C              
#>  [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
#>  [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
#>  [9] LC_ADDRESS=C               LC_TELEPHONE=C            
#> [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       
#> 
#> time zone: America/New_York
#> tzcode source: system (glibc)
#> 
#> attached base packages:
#> [1] stats     graphics  grDevices utils     datasets  methods   base     
#> 
#> other attached packages:
#> [1] biomaRt_2.61.3   tidysbml_0.99.10
#> 
#> loaded via a namespace (and not attached):
#>  [1] KEGGREST_1.45.1         xfun_0.47               bslib_0.8.0            
#>  [4] httr2_1.0.4             Biobase_2.65.1          vctrs_0.6.5            
#>  [7] tools_4.4.1             generics_0.1.3          stats4_4.4.1           
#> [10] curl_5.2.3              tibble_3.2.1            fansi_1.0.6            
#> [13] AnnotationDbi_1.67.0    RSQLite_2.3.7           blob_1.2.4             
#> [16] pkgconfig_2.0.3         dbplyr_2.5.0            S4Vectors_0.43.2       
#> [19] lifecycle_1.0.4         GenomeInfoDbData_1.2.12 compiler_4.4.1         
#> [22] stringr_1.5.1           Biostrings_2.73.1       progress_1.2.3         
#> [25] GenomeInfoDb_1.41.1     htmltools_0.5.8.1       sass_0.4.9             
#> [28] yaml_2.3.10             pillar_1.9.0            crayon_1.5.3           
#> [31] jquerylib_0.1.4         cachem_1.1.0            tidyselect_1.2.1       
#> [34] digest_0.6.37           stringi_1.8.4           dplyr_1.1.4            
#> [37] purrr_1.0.2             fastmap_1.2.0           cli_3.6.3              
#> [40] magrittr_2.0.3          utf8_1.2.4              withr_3.0.1            
#> [43] prettyunits_1.2.0       filelock_1.0.3          UCSC.utils_1.1.0       
#> [46] rappdirs_0.3.3          bit64_4.5.2             rmarkdown_2.28         
#> [49] XVector_0.45.0          httr_1.4.7              bit_4.5.0              
#> [52] png_0.1-8               hms_1.1.3               memoise_2.0.1          
#> [55] evaluate_1.0.0          knitr_1.48              IRanges_2.39.2         
#> [58] BiocFileCache_2.13.0    rlang_1.1.4             glue_1.7.0             
#> [61] DBI_1.2.3               xml2_1.3.6              BiocGenerics_0.51.1    
#> [64] jsonlite_1.8.9          R6_2.5.1                zlibbioc_1.51.1

References

Durinck, Steffen, Yves Moreau, Arek Kasprzyk, Sean Davis, Bart De Moor, Alvis Brazma, and Wolfgang Huber. 2005. “BioMart and Bioconductor: A Powerful Link Between Biological Databases and Microarray Data Analysis.” Bioinformatics 21 (16): 3439–40. https://doi.org/10.1093/bioinformatics/bti525.

Gustavsen, Julia A., Shraddha Pai, Ruth Isserlin, Barry Demchak, and Alexander R. Pico. 2019. “RCy3: Network Biology Using Cytoscape from Within R.” F1000Research. https://doi.org/10.12688/f1000research.20887.3.

Hucka, Michael, Frank T. Bergmann, Claudine Chaouiya, Andreas Dräger, Stefan Hoops, Sarah M. Keating, Matthias König, et al. 2019. “The Systems Biology Markup Language (SBML): Language Specification for Level 3 Version 2 Core Release 2.” Journal of Integrative Bioinformatics 16 (2): 20190021. https://doi.org/doi:10.1515/jib-2019-0021.

Jassal, Bijay. 2016. “Aryl Hydrocarbon Receptor Signalling.” https://reactome.org/content/detail/R-HSA-8937144.

Milacic, Marija, Deidre Beavers, Patrick Conley, Chuqiao Gong, Marc Gillespie, Johannes Griss, Robin Haw, et al. 2023. “The Reactome Pathway Knowledgebase 2024.” Nucleic Acids Research 52 (D1): D672–D678. https://doi.org/10.1093/nar/gkad1025.

Wickham, Hadley, Jim Hester, and Jeroen Ooms. 2023. Xml2: Parse XML. https://xml2.r-lib.org/.