get_multimir: Get microRNA-target Interactions from the multiMiR Package

In multiMiR: Integration of multiple microRNA-target databases with their disease and drug associations

Description

The main function to retrieve predicted and validated miRNA-target interactions and their disease and drug associations from the multiMiR package.

Usage

get_multimir(url = NULL, org = "hsa", mirna = NULL, target = NULL,
  disease.drug = NULL, table = "validated", predicted.cutoff = NULL,
  predicted.cutoff.type = "p", predicted.site = "conserved",
  summary = FALSE, add.link = FALSE, use.tibble = FALSE, limit = NULL,
  legacy.out = FALSE)

get.multimir(url = NULL, org = "hsa", mirna = NULL, target = NULL,
  disease.drug = NULL, table = "validated", predicted.cutoff = NULL,
  predicted.cutoff.type = "p", predicted.site = "conserved",
  summary = FALSE, add.link = FALSE, use.tibble = FALSE, limit = NULL)

Arguments

url	Deprecated. The URL for queries is now defined by the package options multimir.url and multimir.queries.
org	a character string for the organism. Three organisms are supported so far: human ("hsa" (default), "human", or "Homo Sapiens"), mouse ("mmu", "mouse", or "Mus musculus"), and rat ("rno", "rat", or "Rattus norvegicus"). The organism is case insensitive.
mirna	'NULL' (default) or a character string or character vector for the mature miRNA(s). It can be the mature miRNA accession number (i.e. "MIMAT0000072"), mature miRNA ID (i.e. "hsa-miR-199a-3p"), or a combination of both (i.e. c("MIMAT0000065", "hsa-miR-30a-5p")). The character is case insensitive. *See note about the length of list supported.
target	'NULL' (default) or a character string or character vector for the target gene(s). It can be the gene symbol (i.e. c("TP53", "KRAS")), Entrez gene ID (i.e. c(578, 3845)), Ensembl gene ID (i.e. "ENSG00000171791"), or a combination of any of these identifiers (i.e. c("TP53", 3845, "ENSG00000171791")). The character is case insensitive. *See note about the length of list supported.
disease.drug	'NULL' (default) or a character string or character vector for the disease(s) and/or drug(s) (i.e. c("bladder cancer", "cisplatin")). The character is case insensitive.
table	a character string indicating which table(s) in multiMiR to search. Each table contains data from an external database. Options include "validated" (default, to search all validated tables "mirecords", "mirtarbase", and "tarbase"), "predicted" (to search all predicted tables "diana_microt", "elmmo", "microcosm", "miranda", "mirdb", "pictar", "pita", and "targetscan"), "disease.drug" (to search all disease/drug tables "mir2disease", "pharmaco_mir", and "phenomir"), "all" (to search all of the tables above), or an individual table from above.
predicted.cutoff	'NULL' (default) or an integer giving a prediction score cutoff. By default ('NULL'), the cutoff is '20' (search the top 20% if predicted.cutoff.type="p") or '300000' (search the top 300000 (or all records if total < 300000) if predicted.cutoff.type="n").
predicted.cutoff.type	a character indicating the type of prediction score cutoff. This must be either "p" (default, percentage cutoff) or "n" (number cutoff).
predicted.site	a character string indicating the type of predicted target sites to search. This can be one of the strings "conserved", "nonconserved", or "all", and can be abbreviated. This only applies to three of the predicted tables ("miranda", "pita", and "targetscan") that have conservation information of the target sites.
summary	logical. Whether to summarize the result (default = FALSE).
add.link	logical. Whether to add link to external database for each result entry.
use.tibble	logical. Whether to use the data_frame class from the tibble package for returned dataframes. The key benefit for large datasets is more restrictive printing to the console (first 10 rows and only the number of columns that will fit getOption('width')). See ?tible::data_frame for more information.
limit	a positive integer. Limits the number of records returned from each table. Useful in testing potentially large queries.
legacy.out	logical. Whether to return the Bioconductor compatible S4 object or the legacy S3 object (default=FALSE).

Details

get.multimir() has been deprecated and replaced with the get_multimir() version.

get_multimir is the main and recommended function to retrieve information from the multiMiR package. Input to the function must contain at least one of the followings: miRNA(s), target gene(s), and disease and drug term(s).

The setting of predicted.site is applicable to three ("miranda", "pita", and "targetscan") of the eight predicted tables. If predicted.site is "conserved", the function will search conserved target sites annotated by TargetScan, target sites with conservation scores greater than or equal to 0.57 (in human and rat; or 0.566 in mouse) in miRanda, and/or sites with conservation scores greater than or equal to 0.9 in PITA.

Although the summary (if summary=TRUE) can be used to find results that are recorded by combinations of different databases, please note that for predicted interactions a combination approach may not be as effective as a single algorithm because of age or quality of the tool.

Note: The length of the list supported has been increased from version1.0.1. The size is now limited to 20MB which should accommodate most requests. There is a possibility for technical reasons that the query could fail even if the list is under this limit. If this occurs it is recommended that you break up the list into smaller batches and submit them sequentially.

Value

get_multimir returns an S4 object (see ?mmquery_bioc-class containing the queried data and associated metadata. With legacy.out=FALSE (default), the data is a single dataset with association/interaction type defined by the type variable. With legacy.out=TRUE the original S3 object with 3 separate data frames ('predicted', 'validated', and 'disease_drug') is returned.

Examples

  ## search 'hsa-miR-18a-3p' in validated interactions in human
  example1 <- get_multimir(mirna='hsa-miR-18a-3p', summary=TRUE)
  columns(example1)
  ## target genes that are validated by Luciferase assay
  lucif <- select(example1, keytype = "type", keys = "validated", 
                  columns = columns(example1))
  lucif[grep("Luciferase", lucif$experiment), ]
  example1@summary[example1@summary[,"target_symbol"] == "KRAS",]

  ## search 'cisplatin' in disease and drug tables in human
  example2 <- get_multimir(disease.drug='cisplatin', table='disease.drug')
  nrow(example2@data)
  head(example2@data)

multiMiR documentation built on Nov. 8, 2020, 5:46 p.m.