| Title: | miRNAtap: microRNA Targets - Aggregated Predictions |
|---|---|
| Description: | The package facilitates implementation of workflows requiring miRNA predictions, it allows to integrate ranked miRNA target predictions from multiple sources available online and aggregate them with various methods which improves quality of predictions above any of the single sources. Currently predictions are available for Homo sapiens, Mus musculus and Rattus norvegicus (the last one through homology translation). |
| Authors: | Maciej Pajak, T. Ian Simpson |
| Maintainer: | T. Ian Simpson <[email protected]> |
| License: | GPL-2 |
| Version: | 1.41.0 |
| Built: | 2024-10-30 08:49:08 UTC |
| Source: | https://github.com/bioc/miRNAtap |
This function performs aggregation phase of target
prediction for getPredictedTargets.
Consensus ranking is derived from multiple individual rankings.
Available methods include minimum, maximum and geometric mean with further
tuning parameters which promote true positives at the top of the final
ranking
aggregateRanks(ranks, n_valid_srcs, min_src, method = "geom", promote = TRUE)aggregateRanks(ranks, n_valid_srcs, min_src, method = "geom", promote = TRUE)
ranks |
|
n_valid_srcs |
number of valid sources in the dataset |
min_src |
minimum acceptable number fo sources |
method |
|
promote |
add weights to improve accuracy of the method, default
|
data.frame object with ranks per source and aggregate ranks
Maciej Pajak [email protected]
data = data.frame(GeneID=c("15364", "56520", "57781", "58180", "18035"), source1scores=c(0.9,0.5,0.3,NA,NA), source2scores=c(0.7,NA,0.8,0.6,0.5), source3scores=c(0.5,NA,0.3,0.1,0.2)) data #dataframe with scores aggregateRanks(data, n_valid_srcs=3, min_src=2, method='geom') #note how gene 56520 is eliminated as it appeared in fewer than 2 sourcesdata = data.frame(GeneID=c("15364", "56520", "57781", "58180", "18035"), source1scores=c(0.9,0.5,0.3,NA,NA), source2scores=c(0.7,NA,0.8,0.6,0.5), source3scores=c(0.5,NA,0.3,0.1,0.2)) data #dataframe with scores aggregateRanks(data, n_valid_srcs=3, min_src=2, method='geom') #note how gene 56520 is eliminated as it appeared in fewer than 2 sources
This method performs aggregation of target lists from multiple
sources. Aggregated list is more accurate than any list from a single
source. Multiple aggregation methods are available.Direct target data from
five sources for Human and Mouse is supplied through miRNAtap.db
package, for Rat targets are derived through homology translations whenever
direct ones are not available.
getPredictedTargets(mirna, sources = c("pictar", "diana", "targetscan", "miranda","mirdb"), species = "mmu", min_src = 2, method = "geom", promote = TRUE, synonyms = TRUE, both_strands = FALSE, ...)getPredictedTargets(mirna, sources = c("pictar", "diana", "targetscan", "miranda","mirdb"), species = "mmu", min_src = 2, method = "geom", promote = TRUE, synonyms = TRUE, both_strands = FALSE, ...)
mirna |
miRNA in a standard format |
sources |
a list of sources to use for aggregation,
default is all five sources, i.e.
|
species |
species in a standard three-letter acronym, |
min_src |
minimum number of sources required for a target to be considered, default 2 |
method |
method of aggregation - choose from |
promote |
add weights to improve accuracy of the method, default TRUE |
synonyms |
when searching for -3p miRNA automatically also searches for miRNA with the same name but ending with * (some databases list -3p miRNA this way) and other way around, similarly for -5p miRNA, default TRUE |
both_strands |
overrides |
... |
any optional arguments |
Tuning min_src parameter is an easy way of prioritising
precision at the top of the list (high values) or total recall (low values).
For the five default input sources, recommended values are 2, 3, or 4.
data.frame object where row names are entrez IDs of target
genes, ranks from individual sources and aggregated rank are shown
in columns.
If no targets are found in any of the sources NULL and a warning
are returned.
Maciej Pajak [email protected]
Agarwal V, Bell GW, Nam J, Bartel DP. Predicting effective microRNA target sites in mammalian mRNAs. eLife, 4:e05005, (2015).
Griffiths-Jones, S., Saini, H. K., van Dongen, S., and Enright, A. J. (2008). miRBase: tools for microRNA genomics. Nucleic acids research, 36(Database issue):D154-8.
Lall, S., Grun, D., Krek, A., Chen, K., Wang, Y.-L., Dewey, C. N., ... Rajewsky, N. (2006). A genome-wide map of conserved microRNA targets in C. elegans. Current biology : CB, 16(5):460-71.
Paraskevopoulou MD, Georgakilas G, Kostoulas N, Vlachos IS, Vergoulis T, Reczko M, Filippidis C, Dalamagas T, Hatzigeorgiou AG., "DIANA-microT web server v5.0: service integration into miRNA functional analysis workflows.", Nucleic Acids Res. 2013 Jul;41(Web Server issue):W169-73.
Wong N and Wang X (2015) miRDB: an online resource for microRNA target prediction and functional annotations. Nucleic Acids Research. 43(D1):D146-152.
targets <- getPredictedTargets('let-7a',species='hsa', method = 'min') head(targets) #top of the list with minimum aggregation targets2 <- getPredictedTargets('let-7a',species='hsa', method='geom') head(targets2) #top of the list with geometric mean aggregationtargets <- getPredictedTargets('let-7a',species='hsa', method = 'min') head(targets) #top of the list with minimum aggregation targets2 <- getPredictedTargets('let-7a',species='hsa', method='geom') head(targets2) #top of the list with geometric mean aggregation
This function queries precompiled annotation SQLite database which contains miRNA - target gene associations with their respective scores.
getTargetsFromSource(mirna, species = "mmu", source = "diana", synonyms = TRUE, both_strands = FALSE)getTargetsFromSource(mirna, species = "mmu", source = "diana", synonyms = TRUE, both_strands = FALSE)
mirna |
miRNA in a standard format |
species |
species in a standard three-letter acronym, default
|
source |
a source target prediction algorithm table to query, default
|
synonyms |
when searching for -3p miRNA automatically also searches for miRNA with the same name but ending with * (some databases list -3p miRNA this way) and other way around, similarly for -5p miRNA, default TRUE |
both_strands |
overrides |
data.frame object with entrez IDs of target genes and their
scores, if there are no targets found for a given miRNA in a given
table then an empty
Maciej Pajak [email protected]
Friedman, R. C., Farh, K. K.-H., Burge, C. B., and Bartel, D. P. (2009). Most mammalian mRNAs are conserved targets of microRNAs. Genome research, 19(1):92-105.
Griffiths-Jones, S., Saini, H. K., van Dongen, S., and Enright, A. J. (2008). miRBase: tools for microRNA genomics. Nucleic acids research, 36(Database issue):D154-8.
Lall, S., Grun, D., Krek, A., Chen, K., Wang, Y.-L., Dewey, C. N., ... Rajewsky, N. (2006). A genome-wide map of conserved microRNA targets in C. elegans. Current biology : CB, 16(5):460-71.
Maragkakis, M., Vergoulis, T., Alexiou, P., Reczko, M., Plomaritou, K., Gousis, M., ... Hatzigeorgiou, A. G. (2011). DIANA-microT Web server upgrade supports Fly and Worm miRNA target prediction and bibliographic miRNA to disease association. Nucleic Acids Research, 39(Web Server issue), W145-8.
targets <- getTargetsFromSource('let-7a', species='hsa', source='targetscan') head(targets) #top of the listof human targets of let-7a from TargetScan onlytargets <- getTargetsFromSource('let-7a', species='hsa', source='targetscan') head(targets) #top of the listof human targets of let-7a from TargetScan only
object of MirnaDb class holds the sqlite database
connection, and extends AnnotationDb class from AnnotationDbi
package. columns, keys, keytypes and select
methods allow access to database tables and retrieval of miRNA target
information.
select is the most important method, allows querying the
database for predictions from a specific source and species for a
given miRNA
columns(x) keytypes(x) keys(x, keytype, ...) select(x, keys, columns, keytype, ...) ## S4 method for signature 'MirnaDb' columns(x) ## S4 method for signature 'MirnaDb' keytypes(x) ## S4 method for signature 'MirnaDb' keys(x, keytype, ...) ## S4 method for signature 'MirnaDb' select(x, keys, columns, keytype, ...)columns(x) keytypes(x) keys(x, keytype, ...) select(x, keys, columns, keytype, ...) ## S4 method for signature 'MirnaDb' columns(x) ## S4 method for signature 'MirnaDb' keytypes(x) ## S4 method for signature 'MirnaDb' keys(x, keytype, ...) ## S4 method for signature 'MirnaDb' select(x, keys, columns, keytype, ...)
x |
the |
keytype |
the keytype that matches the keys used; the table in which the search should be performed. |
... |
any optional arguments |
keys |
the key to select records for from the database - miRNA name;
all possible keys (miRNAs) are returned by using the |
columns |
in this case same as |
string vectors, for select a data.frame with target
genes and scores
Maciej Pajak [email protected]
#first load the annotations require(miRNAtap.db) #see all available tables keytypes(miRNAtap.db)#first load the annotations require(miRNAtap.db) #see all available tables keytypes(miRNAtap.db)
It is a package with tools to facilitate implementation of workflows requiring miRNA prediction through access to multiple prediction results (DIANA, Targetscan, PicTar, Miranda, and miRDB) and their aggregation. Three aggregation methods are available: minimum, maximum and geometric mean, additional parameters provide further tuning of the results. Predictions are available for Homo sapiens, Mus musculus and Rattus norvegicus (the last one through homology translation).
Maciej Pajak [email protected], Ian Simpson
#direct targets in mouse aggregated from all sources: targets_mouse <- getPredictedTargets('let-7a',species='mmu', method='geom') #homology-translated targets in rat aggregated from all sources targets_rat <- getPredictedTargets('let-7a',species='rno', method='geom')#direct targets in mouse aggregated from all sources: targets_mouse <- getPredictedTargets('let-7a',species='mmu', method='geom') #homology-translated targets in rat aggregated from all sources targets_rat <- getPredictedTargets('let-7a',species='rno', method='geom')
This function maps gene entrez ID between species using homology information from Homologene.
translate(entrezes, from = "mmu", to = "rno", ...)translate(entrezes, from = "mmu", to = "rno", ...)
entrezes |
data.frame with entrez Gene IDs and their scores |
from |
origin species, default |
to |
target species, default |
... |
any optional arguments |
data.frame object with orthologous genes' entrez IDs and corresponding scores
Maciej Pajak [email protected]
mouse_genes <- data.frame(GeneID = c("15364", "56520", "57781", "58180", "18035", "239857")) translate(mouse_genes, from='mmu', to='rno')mouse_genes <- data.frame(GeneID = c("15364", "56520", "57781", "58180", "18035", "239857")) translate(mouse_genes, from='mmu', to='rno')