Package 'STRINGdb'

Description

Take in input a dataframe containing a logFC column that reports the logarithm of the difference in expression level. Add a "color" column to the data frame such that strongly downregulated genes are colored in green and strong upregulated genes are in red. When the down or up-regulation is instead weak the intensity of the color gets weaker as well, accordingly.

Usage

## S4 method for signature 'STRINGdb'
add_diff_exp_color(screen, logFcColStr="logFC" )

Arguments

Value

vector containing the colors

Author(s)

Andrea Franceschini

Description

Add description coluns to the proteins that are present in the data frame given in input. The data frame must contain a column named "STRING_id".

Usage

## S4 method for signature 'STRINGdb'
add_proteins_description(screen)

Arguments

Value

returns the same dataframe given in input with an additional columns containing a description of the proteins.

Author(s)

Andrea Franceschini

Description

coefficient of variation

Usage

coeffOfvar(x)

Arguments

Details

coefficient of variation

Value

coefficient of variation

Author(s)

Andrea Franceschini

Description

delete a column in the data frame

Usage

delColDf(df, colName)

Arguments

Value

data frame

Author(s)

Andrea Franceschini

Description

example of microarray data (data processed from GEO GSE9008)

Usage

data(diff_exp_example1)

Format

Data frames with 20861 observations on the following 3 variables.

gene

a character vector

pvalue

a numeric vector

logFC

a numeric vector

Source

Whyte L, Huang YY, Torres K, Mehta RG. Molecular mechanisms of resveratrol action in lung cancer cells using dual protein and microarray analyses. Cancer Res 2007.

Description

download a file only if it is not present.

Usage

downloadAbsentFile(urlStr, oD = tempdir())

Arguments

Author(s)

Andrea Franceschini

Description

download a STRING file only if it is not present or if it is corrupted.

Usage

downloadAbsentFileSTRING(urlStr, oD = tempdir())

Arguments

Author(s)

Andrea Franceschini

Description

Loads and returns STRING aliases. Depending on takeFirst, this returns either all alias mappings or a single preferred mapping for ambiguous aliases.

Usage

## S4 method for signature 'STRINGdb'
get_aliases(takeFirst=TRUE, usePreferredSources=TRUE)

Arguments

Value

a data frame containing STRING aliases. With takeFirst=FALSE, all alias mappings are returned. With takeFirst=TRUE, ambiguous aliases are collapsed to one STRING identifier.

Author(s)

Andrea Franceschini

Description

Loads and returns STRING annotations (i.e. GO annotations, KEGG pathways, domain databases). The annotations are stored in the "annotations" variable.

Usage

## S4 method for signature 'STRINGdb'
get_annotations( )

Value

a data frame containing the annotations to the STRING proteins (e.g. GeneOntology, KEGG pathways, InterPro domains)

Author(s)

Andrea Franceschini

Description

Returns a data frame with the description of every STRING annotation term (it downloads and caches the information the first time that is called).

Usage

## S4 method for signature 'STRINGdb'
get_annotations_desc()

Value

data frame with the description of every STRING annotation term.

Author(s)

Andrea Franceschini

Description

Returns the interaction graph as an object of the graph package in Bioconductor. For network_type="regulatory", the returned graphNEL object preserves edge direction.

Usage

## S4 method for signature 'STRINGdb'
get_bioc_graph()

Value

interaction graph as an object of the graph package in Bioconductor.

Author(s)

Andrea Franceschini

Description

Returns a list of clusters of interacting proteins. Clustering is not available for the directed network_type="regulatory" network. See the iGraph (http://igraph.sourceforge.net/) documentation for additional information on the algorithms.

Usage

## S4 method for signature 'STRINGdb'
get_clusters(string_ids, algorithm="fastgreedy")

Arguments

Value

list of clusters of interacting proteins.

Author(s)

Andrea Franceschini

Description

Returns the enrichment in pathways of the vector of STRING proteins that is given in input.

Usage

## S4 method for signature 'STRINGdb'
get_enrichment(string_ids, category = "All", methodMT = NULL, iea = NULL, minScore = NULL, max_fdr = NULL, min_gene_count = NULL, min_strength = NULL, min_signal = NULL, merge_similarity = NULL)

Arguments

Details

STRING applies the optional filters like the filters of the enrichment table on its website. Without them (the default NULL), no term is filtered. The returned table has no strength and signal columns.

The filters need STRING version 12.5 or later. An older version, such as version="12.0", would ignore them, so they are not used there, with a warning.

Value

Data frame containing the enrichment in pathways of the vector of STRING proteins that is given in input.

Author(s)

Andrea Franceschini

Description

Returns a STRING enrichment figure for the given identifiers and enrichment category.

Usage

## S4 method for signature 'STRINGdb'
get_enrichment_figure(string_ids, category = "Process", file = NULL, output_format = "image", group_by_similarity = NULL, color_palette = "mint_blue", number_of_term_shown = NULL, x_axis = "signal", caller_identity = "STRINGdb-package", graph_type = "dotplot", sort_by = NULL, max_fdr = NULL, min_gene_count = NULL, min_strength = NULL, min_signal = NULL, merge_similarity = NULL)

Arguments

Details

STRING applies the optional filters like the filters of the enrichment table on its website; without them (the default NULL), no term is filtered. When STRING refuses the request, the method stops with STRING's explanation.

graph_type, sort_by and the filters need STRING version 12.5 or later. An older version, such as version="12.0", would ignore them, so they are not used there, with a warning.

Value

For output_format="image" and output_format="highres_image", returns a PNG image array. For output_format="svg", returns the SVG markup as text.

Author(s)

Damian Szklarczyk

Description

Returns concise primary, secondary, and tertiary descriptions of a protein set using the STRING gene-set description API.

Usage

## S4 method for signature 'STRINGdb'
get_geneset_description(string_ids, caller_identity="STRINGdb-package")

Arguments

Value

A one-row data frame containing primary, secondary, and tertiary descriptions when they are available.

Author(s)

Damian Szklarczyk

Description

Return an igraph object with the STRING network (for information about iGraph visit http://igraph.sourceforge.net). For network_type="regulatory", the graph is directed from regulator to target.

Usage

## S4 method for signature 'STRINGdb'
get_graph()

Value

igraph object with the STRING network

Author(s)

Andrea Franceschini

References

Csardi G, Nepusz T: The igraph software package for complex network research, InterJournal, Complex Systems 1695. 2006. http://igraph.sf.net

See Also

In order to simplify the most common tasks, we do also provide convenient functions that wrap some iGraph functions. get_interactions(string_ids) # returns the interactions in between the input proteins get_neighbors(string_ids) # Get the neighborhoods of a protein (or of a vector of proteins) that is given in input. get_subnetwork(string_ids) # returns a subgraph from the given input proteins

Description

Returns the list of closest homologs (as measured by bitscore) of the given input identifiers in all STRING species or single target species.

Usage

## S4 method for signature 'STRINGdb'
get_homologs_besthits(string_ids, target_species_id=NULL)

Arguments

Value

Data frame containing the best blast hits x species of the given input identifiers.

Author(s)

Andrea Franceschini

Description

Returns the interaction partners of the input proteins using the locally loaded STRING graph. The returned data frame preserves the edge attributes available for the current link_data setting. For a regulatory network, from and to preserve the regulator-to-target direction.

Usage

## S4 method for signature 'STRINGdb'
get_interaction_partners(string_ids, required_score=NULL, limit=NULL)

Arguments

Value

Data frame containing one row per query-partner interaction. The leading columns are from, to, combined_score, from_name and to_name, followed by any additional edge attributes loaded for the current link_data mode. For functional and physical networks, from is the queried protein and to is its interaction partner. For regulatory networks, they are the original regulator and target. If limit is used, rows are ordered by combined_score.

Author(s)

Damian Szklarczyk

Description

Shows the interactions in between the proteins that are given in input. For a regulatory network, from is the regulator and to is the regulated protein.

Usage

## S4 method for signature 'STRINGdb'
get_interactions(string_ids)

Arguments

Value

Data frame containing the interactions in between the input proteins. The leading columns are from, to, combined_score, from_name and to_name, followed by any additional loaded edge attributes.

Author(s)

Damian Szklarczyk

Description

Returns a short link to the network page of the STRING website that shows the protein interactions between the given identifiers.

Usage

## S4 method for signature 'STRINGdb'
get_link(string_ids=NULL, required_score=NULL, network_flavor="evidence", payload_id=NULL, network_term_id=NULL, hide_node_labels=NULL, hide_disconnected_nodes=NULL, block_structure_pics_in_bubbles=NULL, flat_node_design=TRUE, center_node_labels=NULL, custom_label_font_size=NULL, caller_identity="STRINGdb-package", typed_physical_edges=NULL, typed_regulatory_edges=NULL, show_regulatory_signs=NULL)

Arguments

Value

short link to the network page of the STRING website that shows the protein interactions between the input identifiers.

Author(s)

Damian Szklarczyk

Description

Get the neighborhoods of a protein (or of a vector of proteins) that is given in input.

Usage

## S4 method for signature 'STRINGdb'
get_neighbors(string_ids, mode="all")

Arguments

Value

vector containing the neighborhoods of a protein (or of a vector of proteins) that is given in input.

Author(s)

Andrea Franceschini

Description

Returns the list of paralogs of the given input in their species.

Usage

## S4 method for signature 'STRINGdb'
get_paralogs(string_ids)

Arguments

Value

Data frame containing the within-species homology hits of the input identifiers, with columns ncbiTaxonId_A, stringId_A, ncbiTaxonId_B, stringId_B, and bitscore.

Author(s)

Andrea Franceschini

Description

Returns a STRING network image for the given identifiers or for a STRING functional term.

Usage

## S4 method for signature 'STRINGdb'
get_png(string_ids=NULL, required_score=NULL, network_flavor="evidence", file=NULL, payload_id=NULL, output_format="image", network_term_id=NULL, hide_node_labels=NULL, hide_disconnected_nodes=NULL, block_structure_pics_in_bubbles=NULL, flat_node_design=TRUE, center_node_labels=NULL, custom_label_font_size=NULL, caller_identity="STRINGdb-package", typed_physical_edges=NULL, typed_regulatory_edges=NULL, show_regulatory_signs=NULL, network_clustering_algorithm=NULL, network_clustering_parameter=NULL, cluster_aware_layout=NULL, network_clustering_edge_visibility=NULL, show_cluster_names=NULL, cluster_name_size=NULL)

Arguments

Details

A network of more than 1,000 proteins is drawn simplified, as on the STRING website: with confidence edges (whatever network_flavor says) and without structure pictures in the nodes.

STRING draws a PNG (output_format="image" or "highres_image") of at most 150,000 interactions. For a larger network, and for any other request that STRING refuses, the method stops with STRING's explanation; for the PNG, ask for the SVG (output_format="svg") or use a higher required_score.

The clustering options need STRING version 12.5 or later. An older version, such as version="12.0", would ignore them, so they are not used there, with a warning.

Value

For output_format="image" and output_format="highres_image", returns a PNG image array. For output_format="svg", returns the SVG markup as text.

Author(s)

Damian Szklarczyk

Description

Returns a pvalue representing the enrichment in interactions of the list of proteins (i.e. the probability to obtain such a number of interactions by chance). The calculation uses the object's selected functional, physical, or regulatory network type.

Usage

## S4 method for signature 'STRINGdb'
get_ppi_enrichment(string_ids)

Arguments

Value

Returns a pvalue representing the enrichment in interactions of the list of proteins (i.e. the probability to obtain such a number of interactions by chance).

Author(s)

Andrea Franceschini

Description

Returns the STRING proteins data frame. (it downloads and caches the information the first time that is called).

Usage

## S4 method for signature 'STRINGdb'
get_proteins()

Value

STRING proteins data frame.

Author(s)

Andrea Franceschini

Description

Returns the subgraph generated by the given input proteins. For a regulatory network, regulator-to-target edge directions are preserved.

Usage

## S4 method for signature 'STRINGdb'
get_subnetwork(string_ids )

Arguments

Value

Returns the subgraph (i.e. an iGraph object) generated by the given input proteins.

Author(s)

Andrea Franceschini

Description

Returns a summary of the STRING sub-network containing the identifiers provided in input. The summary and its interaction-enrichment p-value use the object's selected network type.

Usage

## S4 method for signature 'STRINGdb'
get_summary(string_ids, required_score=NULL)

Arguments

Value

Returns a summary (i.e. a text description) of the STRING sub-network containing the identifiers provided in input.

Author(s)

Andrea Franceschini

Description

Returns the proteins annotated to belong to a given term.

Usage

## S4 method for signature 'STRINGdb'
get_term_proteins(term_ids, string_ids=NULL, enableIEA=TRUE)

Arguments

Value

Returns the proteins annotated to belong to a given term.

Author(s)

Andrea Franceschini

Description

example of a sorted list of protein-protein interactions, resulta our cooccurrence algorithm (SVD_Phy)

Usage

data(interactions_example)

Format

Data frames with 20861 observations on the following 3 variables.

proteinA

a character vector

proteinB

a character vector

score

a numeric vector

Description

Downloads and returns the STRING network (the network is set also in the graph variable of the STRING_db object). When possible, the download uses the threshold-specific streamed network file matching the current score_threshold. For network_type="regulatory", the returned igraph object is directed: an edge from protein1 to protein2 represents regulation of the latter by the former.

It makes use of the variables: "backgroundV" vector containing STRING identifiers to be used as background (i.e. the STRING network loaded will contain only the proteins that are present also in this vector) "score_threshold" STRING combined score threshold (the network loaded contains only interactions having a combined score greater than this threshold)

Usage

## S4 method for signature 'STRINGdb'
load()

Value

STRING network (i.e. an iGraph object. For info look to http://igraph.sourceforge.net)

Author(s)

Andrea Franceschini and Damian Szklarczyk

Description

Force download and loading of all the files (so that you can later store the object on the hard disk if you like). It makes use of the variables: "backgroundV" vector containing STRING identifiers to be used as background (i.e. the STRING network loaded will contain only the proteins that are present also in this vector) "score_threshold" STRING combined score threshold (the network loaded contains only interactions having a combined score greater than this threshold)

Usage

## S4 method for signature 'STRINGdb'
load_all()

Author(s)

Andrea Franceschini

Description

Maps the gene identifiers of the input dataframe to STRING identifiers. It returns the input dataframe with the "STRING_id" additional column.

Usage

## S4 method for signature 'STRINGdb'
map(my_data_frame, my_data_frame_id_col_names, takeFirst=TRUE, removeUnmappedRows=FALSE, quiet=FALSE, usePreferredSources=TRUE)

Arguments

Value

Returns the dataframe that is given in input with the "STRING_id" additional column. The first column of my_data_frame_id_col_names comes first, followed by the other columns of the input and by "STRING_id". The identifiers are compared without regard to case. An NA or empty identifier is never mapped (its STRING_id is NA). The rows keep the order of the input, also when several identifier columns are used, and each row keeps its own identifiers as given; with takeFirst=FALSE, a row whose identifier matches several STRING proteins is repeated, once for each of them.

Author(s)

Andrea Franceschini

Description

Maps the gene identifiers of the input vector to STRING identifiers (using a take first approach). It returns a vector with the STRING identifiers of the mapped proteins.

Usage

## S4 method for signature 'STRINGdb'
mp(protein_aliases)

Arguments

Value

It returns a vector with the STRING identifiers of the mapped proteins.

Author(s)

Andrea Franceschini

Description

mapping function (it add the possibility to map using more than one column of the data frame)

Usage

multi_map_df(dfToMap, dfMap, strColsFrom, strColFromDfMap, strColToDfMap, caseSensitive=FALSE)

Arguments

Value

data frame with an additional column containing the result of the mapping. The first column of strColsFrom comes first, followed by the other columns of dfToMap and by the result column. An NA or empty identifier is never mapped. The rows keep the order of dfToMap, also when several columns are used, and each row keeps its own identifiers as given; a row whose identifier maps to several targets is repeated, once for each of them.

Author(s)

Andrea Franceschini

Description

Plots an image of the STRING network with the given proteins.

Usage

## S4 method for signature 'STRINGdb'
plot_network(string_ids, payload_id=NULL, required_score=NULL, add_link=FALSE, network_flavor="evidence", add_summary=TRUE, typed_physical_edges=NULL, typed_regulatory_edges=NULL, show_regulatory_signs=NULL)

Arguments

Details

A network of more than 1,000 proteins is drawn simplified, as on the STRING website: with confidence edges (whatever network_flavor says) and without structure pictures in the nodes.

The picture is a PNG, which STRING draws for at most 150,000 interactions. For a larger network the method stops with STRING's explanation: use a higher required_score, or get_png with output_format="svg".

Author(s)

Andrea Franceschini

Description

Posts the input to STRING and returns an identifier that you can use to access the payload when you enter in our website.

Usage

## S4 method for signature 'STRINGdb'
post_payload(stringIds, colors=NULL, comments=NULL, links=NULL, iframe_urls=NULL, logo_imgF=NULL, legend_imgF=NULL )

Arguments

Value

identifier of the payload.

Author(s)

Andrea Franceschini

Description

With this method it is possible to remove the interactions that are composed by a pair of homologous/similar proteins, having a similarity bitscore between each other higher than a threshold.

Usage

## S4 method for signature 'STRINGdb'
remove_homologous_interactions(interactions_dataframe, bitscore_threshold = 60)

Arguments

Value

interactions data frame where the homologous pairs have been removed, from the input interactions' data frame

Author(s)

Andrea Franceschini

Description

Rename a column of a data frame

Usage

renameColDf(df, colOldName, colNewName)

Arguments

Value

data frame with the column name changed

Author(s)

Andrea Franceschini

Description

With this method you can specify a vector of proteins to be used as background. The network is reloaded and only the proteins that are present in the background vector are inserted in the graph. Besides, the background is taken in consideration for all the enrichment statistics. If you already created a STRINGdb object, calling set_background is sufficient and you do not need to instantiate a new object again.

Usage

## S4 method for signature 'STRINGdb'
set_background(background_vector )

Arguments

Author(s)

Andrea Franceschini

Description

The R package STRINGdb provides a convenient interface to STRING, a protein-protein interaction database and functional enrichment analysis tool, for R/Bioconductor users. Please look at the manual/vignette for additional information and examples on how to use the package. STRING is a database of known and predicted protein-protein interactions. It contains information from numerous sources, including experimental repositories, computational prediction methods, and public text collections. Each interaction is associated with a combined confidence score that integrates the different evidence channels. STRING v12.5 contains information on 59.3 million proteins from 12,535 organisms and more than 28 billion interactions. It provides functional, physical, and directed regulatory networks. The STRING web interface is freely accessible at: https://string-db.org/

Extends

All reference classes extend and inherit methods from "envRefClass".

Fields

annotations:

Object of class data.frame ~~

annotations_description:

Object of class data.frame ~~

graph:

Object of class igraph ~~

proteins:

Object of class data.frame ~~

speciesList:

Object of class data.frame ~~

species:

Object of class numeric ~~

version:

Object of class character. The STRING version, e.g. "12.5"; a version without a dot gets ".0" ("11" means "11.0"). For the versions known to the package (11.0, 11.0b, 11.5, 12.0 and 12.5) the object is created without a request to STRING, so it can also be created offline. Without a version, STRING is asked for its current version; when STRING cannot be reached within a few seconds, the newest known version is used and a warning is printed. Any other version is looked up on STRING, and creating the object stops with an error when that version is not available or STRING cannot be reached. ~~

network_type:

Object of class character. One of "full", "physical", or "regulatory". The regulatory network requires STRING v12.5 or later. ~~

link_data:

Object of class character. One of "combined_only", "detailed", or "full". ~~

input_directory:

Object of class character ~~

backgroundV:

Object of class vector ~~

score_threshold:

Object of class numeric ~~

Methods

set_background(background_vector):

~~

post_payload(stringIds, colors, comments, links, iframe_urls, logo_imgF, legend_imgF):

~~

plot_network(string_ids, payload_id, required_score):

~~

plot_ppi_enrichment(string_ids, file, sliceWindow, edgeWindow, windowExtendedReferenceThreshold, minVal, title):

~~

map(my_data_frame, my_data_frame_id_col_names, takeFirst, removeUnmappedRows, quiet):

~~

load():

~~

get_term_proteins(term_ids, string_ids, enableIEA):

~~

get_summary(string_ids):

~~

get_subnetwork(string_ids):

~~

get_ppi_enrichment_full(string_ids, sliceWindow, edgeWindow, windowExtendedReferenceThreshold, growingWindowLimit):

~~

get_ppi_enrichment(string_ids):

~~

get_proteins():

~~

get_enrichment_figure(string_ids, category="Process", file, output_format="image", group_by_similarity, color_palette="mint_blue", number_of_term_shown, x_axis="signal", caller_identity="STRINGdb-package", graph_type="dotplot", sort_by, max_fdr, min_gene_count, min_strength, min_signal, merge_similarity):

~~

get_png(string_ids, required_score, network_flavor, file, payload_id, output_format, network_term_id, hide_node_labels, hide_disconnected_nodes, block_structure_pics_in_bubbles, flat_node_design=TRUE, center_node_labels, custom_label_font_size, caller_identity, typed_physical_edges, typed_regulatory_edges, show_regulatory_signs, network_clustering_algorithm, network_clustering_parameter, cluster_aware_layout, network_clustering_edge_visibility, show_cluster_names, cluster_name_size):

~~

get_neighbors(string_ids, mode="all"):

~~

get_interaction_partners(string_ids, required_score, limit):

~~

get_link(string_ids, required_score, network_flavor, payload_id, network_term_id, hide_node_labels, hide_disconnected_nodes, block_structure_pics_in_bubbles, flat_node_design=TRUE, center_node_labels, custom_label_font_size, caller_identity, typed_physical_edges, typed_regulatory_edges, show_regulatory_signs):

~~

get_interactions(string_ids):

~~

get_homologs_besthits(string_ids, symbets, target_species_id, bitscore_threshold):

~~

get_homologs(string_ids, target_species_id, bitscore_threshold):

~~

get_graph():

~~

get_enrichment(string_ids, category, methodMT, iea, minScore, max_fdr, min_gene_count, min_strength, min_signal, merge_similarity):

~~

get_geneset_description(string_ids, caller_identity="STRINGdb-package"):

~~

get_clusters(string_ids, algorithm):

~~

get_annotations_desc():

~~

get_annotations():

~~

load_all():

~~

initialize(...):

~~

add_proteins_description(screen):

~~

add_diff_exp_color(screen, logFcColStr):

~~

show():

~~

Author(s)

Andrea Franceschini and Damian Szklarczyk

References

Szklarczyk D, Nastou K, Koutrouli M, Kirsch R, Mehryary F, Hachilif R, Hu D, Peluso ME, Huang Q, Fang T, Doncheva NT, Pyysalo S, Bork P, Jensen LJ, von Mering C. The STRING database in 2025: protein networks with directionality of regulation. Nucleic Acids Res. 2025 Jan 6;53(D1):D730-D737. doi: 10.1093/nar/gkae1113.

Examples

showClass("STRINGdb")