miRNA target enrichment

Question

Does anyone know of a package or function that takes in a transcript ID (ENSTXXXXXXXXXX) of a miRNA and one of an mRNA and outputs whether or not that gene is a target for that miRNA?

I have looked around the Bioconductor packages ("mirbase.db", etc) but I cannot find one that does that.

Answer 1

I have just been looking into a similar problem mapping miRNA's to gene targets, and vice versa. One method I found was by using the targetscan.Hs.eg.db (assuming human data here), which can be obtained through bioconductor:

biocLite('targetscan.Hs.eg.db')    

library('org.Hs.eg.db') # required package
library('targetscan.Hs.eg.db')

get(get("SLC2A4", revmap(org.Hs.egSYMBOL)), targetscan.Hs.egTARGETS)
[1] "miR-150/5127"                                                                             
[2] "miR-199ab-5p"                                                                             
[3] "miR-17/17-5p/20ab/20b-5p/93/106ab/427/518a-3p/519d"                                       
[4] "miR-93/93a/105/106a/291a-3p/294/295/302abcde/372/373/428/519a/520be/520acd-3p/1378/1420ac"
[5] "miR-31"   

The above is using Gene_Symbol (there are plenty of ways to convert transcript ID to symbol). This seems to list out the miRNA targets found in the targetscan database for the selected gene, in this case SLC2A4.

Answer 2

You can use targetHub database to get the targets using mirna or gene identifiers.

Although this method doesn't specify whether a given miRNA-gene pair interact, it provides a more flexible way to query your mirna/gene for target information using various prediction algorithms.

Check the following function to access the mir-gene interaction data with any of the prediction methods described in targetHub. Although there is no specific R package provided, you can use simple HTTP calls to access the targetHub database.

library(RJSONIO) # function to extract mirna (mature) target genes
                 # from targetHub using any prediction algorithm

targetHub.byMethods <- function(mir.name, method) {
tarhub.matmirna <- 'http://app1.bioinformatics.mdanderson.org/tarhub/_design/basic/_view/by_matureMIRmethod'
method <- gsub("\\+", "%2B", method)
data.link <- gsub("\\\"", "%22", paste(tarhub.matmirna, '?key=["', mir.name, '","', method ,'"]&include_docs=true', sep=''))
json.data <- paste(readLines(data.link), collapse='') #get data from targetHub
target_data <- fromJSON(json.data) #convert json formatted data to list
target_Info <- NA
if(is.list(target_data) & (length(target_data$rows) > 0)) {
       target_data <- target_data$rows
       target_Info <- matrix(nrow = length(target_data), ncol = 3)
       colnames(target_Info) <- c("Gene_Symbol", "Gene_EntrezID", "miRNA")
       for(i in 1:length(target_data)) {
            target_Info[i,1] = target_data[[i]]$doc$Gene_Symbol
            target_Info[i,2] = target_data[[i]]$doc$Gene_EntrezID
            target_Info[i,3] = target_data[[i]]$doc$miRNA
       }    
    }
    target_Info
}

#usage
miRandaTargets <-  targetHub.byMethods("hsa-miR-200c-3p", "targetscan")
targetscanTargets <-  targetHub.byMethods("hsa-miR-200c-3p", "miranda")
miRanda_targetscan_Targets <-  targetHub.byMethods("hsa-miR-200c-3p", "miranda+targetscan")

To customize the output, you can check the targetHub API link.