R/app.intro.R
In ETHZ-INS/enrichMiR: enrichMiR: miRNA target enrichment analysis
Defines functions
.exampleGeneset
.exampleBackground
.getBrowserCompTable
.getTestsTable
.testDescription
.testIntro
.getHelpModal
.getAppIntro
.getAppIntro <- function(){
data.frame(
element=c(NA, "#menu_species", "#collection_input", "#exprMirs_box",
"#expressed_mirna_outer", "#menu_input", "#input_type",
"#example_dea", "#menu_input", "#menu_enrich", "#enrichbtn_outer",
"#resultsbox", "#menu_cdplot", "#mir_fam", "#cdplot_outer", NA),
intro=c(
"This introduction will walk you through the usage of the enrichMiR app.
<br/><br/>
You can use the Next/Previous buttons to navigate the tutorial, and leave
it at any time by clicking outside it.",
"The first step is to make sure that you're working with the right species
and annotation. To set this, you open the 'Species and miRNAs' tab.",
"Once you've selected a species, you will be able to see and select one
of its available target annotations (see the help button for information
about the annotations). For the sake of this example, leave it at the
starting default, i.e. human 'targetScan conserved miRNA BS'.",
"This is optional, but you can also specify expressed miRNAs by expanding
this box (click the \"+\" on the right).<br/>This will be used to
restrict the miRNAs considered for enrichment analysis. In addition, you
will be able to visualize given expression values in the results.",
"There are three ways to provide this information:<br/>
1) you can specify a list of expressed miRNAs in 'Custom set';<br/>
2) you can upload a table of miRNA expression levels;<br/>
3) you can pick pre-compiled miRNA expression profiles for your
tissue/celltype of interest.<br/><br/>Not that you do not necessarily
need to specify expressed miRNA, but this is likely to give you better
results.",
"Once this is done, the next step is to provide the signal in which
enrichment should be looked for (e.g. your genes of interest, or
differential expression signature). To do this, click on the Input tab.",
"There are two main types of input:<br/>
1) the results of a differential expression analysis (DEA), which you can
upload using the 'Upload DEA results' tab, or<br/>
2) a set of genes to be tested for over-representation (the set can be
specified either manually or by selecting genes belonging to a given GO
category)<br/>
When specifying a set of genes, these will be tested for
over-representation among targets in comparison to a background set of
genes (also known as 'universe'). This is critical to over-representation
analysis! See the corresponding help button for more information about the
choice of a background.<br/><br/>
For the purpose of this example, select the 'Upload DEA results' tab.",
"For this tutorial, we'll use an in-built example DEA, which you can load
by clicking the 'Use example DEA' button.<br/><br/>Once you've done so,
you will see on the right that it successfully read the DEA file.",
"If you've provided a valid input, the input menu on the left will show
a blue badge instead of a red one.",
"We're now ready to launch the enrichment analysis!<br/><br/>First go to
the 'Enrichment analysis' tab",
"Then click the 'Run enrichMir!' button. The analysis will take a moment.",
"Once it's over, an enrichment plot will appear, displaying the
results. You can hover on the points for more information on the top hits.
(For more information on the enrichment plot and its interpretation, click
the help button above it).<br/><br/>
You can also visualize (and download) the results as a table in the
'Results Table' tab.",
"Independently of an enrichment analysis, when your input is a
differential expression analysis (DEA) result you also have the
possibility to create cumulative distribution (CD) plots for predicted
targets of any miRNA. To try this out, open the 'CD plot' tab.",
"By default, a CD plot is created for the first miRNA in the alphabetical
list. To select your miRNA of interest, click in the selector, erase the
current miRNA (using the backspace button), and start typing, for example
'133a'. The miRNA compatible with your selection will show up, and you can
select the one you are interested in. Select a miR-133a-3p, which is a
member of the family whose targets showed the strongest enrichment in
this DEA.",
"After a moment, the cumulative distribution plot will appear, showing
the foldchange distributions of different types of targets and
non-targets. In the presence of a real miRNA effect (that is gene expression
changes were primarily caused by a microRNA), we expect to see a
gradual dose-response pattern (targets with stronger binding sites also
have stronger log-foldchanges), as is the case in this example.<br/><br/>
Note that we could also have gotten to this CD plot by clicking on the
corresponding point in the enrichment plot of the previous tab!",
"This concludes our tutorial. While there are additionnal options here
and there, these are the app's main functionalities."
)
)
}
.getHelpModal <- function(topic){
switch(topic,
overview=modalDialog(title="enrichMiR workflow overview", easyClose=TRUE, size = "l",
tags$img(src="overview.png",height = 525, width = 750,
style="display: block; margin-left: auto; margin-right: auto;"),
tags$head(tags$style("#shiny-modal img { max-width: 100%; }")),
tags$p("For further information regarding the individual steps consult the
specifi help pages or conduct the interactive tour."),
tags$p("This website is free and open to all users and there is no login requirement.")
),
collections=modalDialog(title="Binding sites collections", easyClose=TRUE,
tags$p(
"Each collection represents a set of binding sites (validated or predicted) for
(a set of) miRNAs or families. Not all collections are available in all species.
Here is a brief description of each type of collection:"),
tags$ul(
tags$li(tags$b("TargetScan: "),
"miRNA binding site predictions from ",
tags$a( href="http://www.targetscan.org","TargetScan",
target="_blank"),
" (version 8). These are available in two flavors: using all
sites, or only sites conserved across vertebrates. In both cases,
only one UTR per gene (typically the longest) is considered."),
tags$li(tags$b("ScanMiR: "),
"miRNA binding site predictions using the ",
tags$a(href="https://www.bioconductor.org/packages/devel/bioc/html/scanMiR.html",
"scanMiR", target="_blank"),
"algorithm. ScanMiR employs a transcript-level affinity prediction method and is
compatible with both gene and transcript IDs. To ensure fast computing times, we restricted
the scanMiR annotation to 7mer and 8mer binding sites in the 3'UTR."),
tags$li(tags$b("miRTarBase: "),
"Experimentally-validated miRNA targets from the ",
tags$a(href="https://mirtarbase.cuhk.edu.cn/~miRTarBase/",
"miRTarBase", target="_blank"),
"database. While these targets are more trustworthy, a
sufficient number of targets is available for only a
proportion of the miRNAs, and some targets might have been
missed due to the cellular context of the experiments.
Because this collection does not include individual sites,
only the simplest over-representation test is available."),
tags$li(tags$b("oRNAment: "),
"RNA binding protein (RBPs) bindings from the ",
tags$a(href="http://rnabiology.ircm.qc.ca", "oRNAment",
target="_blank"), " database.", tags$br(),
"Note that enrichMiR was developed for miRNA bindings, and its
usage for RBPs is experimental.")
)
),
deaformat=modalDialog(title="Required format of the DEA table", easyClose=TRUE,
tags$p("The app should recognized the differential expression tables
generated from common RNAseq DEA package and saved as
character-delimited files, assuming that the gene names/IDs
are in the first column (which is normally the case).
In case of problem, try to format your table in this way:"),
tags$pre(
"Gene,logFC,PValue,FDR
CERS2,-1.54,1.37e-19,8.85e-16
NR6A1,-1.76,3.92e-19,1.77e-15
TAGLN2,-1.92,4.47e-17,1.08e-13
FTL,-1.69,1.68e-15,2.05e-12
TPM4,-1.41,1.83e-10,6.86e-08
..."),
tags$p("The character delimiter needs not be commas, and should be
automatically recognized. With most target annotations, both Ensembl
IDs and gene symbols will be recognized, and the scanMiR annotation
additionally supports transcript-level DEAs with Ensembl transcript IDs."),
tags$p("Because the accuracy of the fold-change estimates increases
with the genes' read count, we recommend filtering the DEA table
to include only highly expressed genes (e.g. the top 5000) before
upload.")
),
enrichplot=modalDialog(title="Enrichment plot", easyClose=TRUE, tags$p(
"The enrichment plot shows the significance of the enrichment on the
y-axis, and the magnitude of the enrichment on the x-axis. For basic
overlap/over-representation tests, the enrichment is the fold
increase over the overlap expected by chance. For continuous tests
(based on the results of a differential expression analysis), the
enrichment represents a magnitude of association with the
log-foldchanges of the putative targets."),
tags$p("In general, we recommend regarding the statistical significance
of individual posttranscriptional regulators obtained via the
different enrichment tests rather as rank instead of interpreting
them as finite results. Hence high-ranking miRNAs/families are
suggested to exert a functional role in the gene set at glance,
though certainly need to be further validated."),
tags$p("The size of the
points is proportional to the number of targets with binding sites for
the given miRNA/family. If miRNA expression data was given, the
points are colored by this expression."),
tags$p("Hovering on the points/families will provide extra
information about the family members and the test statistics."),
tags$p("In case a DEA has been provided as input, double-clicking on a
miRNA/family will automatically open the corresponding CD-plot."),
tags$p("The plot size can be manually adapted by dragging the lower right part
with the mouse.")
),
browsercompatibility=modalDialog(title="Browser compatibility", easyClose=TRUE,
fluidRow(column(8, align="center", tableOutput("browsercomp")))
),
cdplot=modalDialog(title="Cumulative distribution (CD) plots", easyClose=TRUE,
tags$p(
"Cumulative distribution plots are used to visualize differences in
log-foldchange distributions between groups of genes/transcripts.
Briefly, the genes in each sets are ranked according to their
foldchange (the x-axis), while the y-axis shows the proportion of
genes in the set that have a foldchange below or equal to that given
on x. In this way, small shifts in foldchanges can easily be
visualized."), tags$p(
"There are different ways of splitting the genes into groups, which
can be selected using the 'Split by' dropdown list. We recommend
splitting by best binding type if this information is available in
the collection you are using, or otherwise splitting by predicted
repression score (if available). This is the default (Automatic)
procedure."),
tags$p(renderPlot({
CDplot(list("no site"=rnorm(600,mean=0,sd=0.8),
"7mer"=c(rnorm(200,0),rnorm(200, mean=-0.5)),
"8mer"=c(rnorm(100,0),rnorm(200, mean=-1))), size=1.3) +
scale_color_manual(values=enrichMiR:::.siteTypeColors()[
c("no site", "7mer", "8mer")]) + xlab("logFC") + xlim(-3,2)
})),
tags$p("In the presence of a real miRNA effect, we expect to see a
gradual dose-response pattern, as shown above, meaning that
targets with stronger binding sites also have stronger
log-foldchanges (i.e. lower logFC, in the case of an increased
miRNA activity, as in this example)."),
tags$p("If your CD plot shows very angular lines, rather than
lines that approach curves, this means that your sets do not contain
enough elements. An example of this is given below:"),
tags$p(renderPlot({
CDplot(list("no site"=rnorm(600,mean=0,sd=0.8),
"7mer"=c(rnorm(8,0),rnorm(15, mean=-0.5)),
"8mer"=c(-2.8, -0.5, 0.5)), size=1.3) +
scale_color_manual(values=enrichMiR:::.siteTypeColors()[
c("no site", "7mer", "8mer")]) + xlab("logFC") + xlim(-3,2)
})),
tags$p("In such cases, if you are using score intervals, try reducing
the number of sets in the plot options. If you are using sites or
site types, this means that the miRNA of interest does not have
enough sites of this type in your dataset. This can especially
happen when using TargetScan ", tags$em("conserved"), "sites (or
miRTarBase sites), in which case you can try using all predicted
sites instead."),
tags$p("Because the accuracy of the fold-change estimates increases
with the genes' read count, we recommend using only highly
expressed genes to plot cummulative distributions of predicted
miRNA bindings, e.g. the top 5'000 - 7'000 expressed genes of a
dataset."),
tags$p("Note that 'no site' indicates genes which are not annotated
to have a binding site for that miRNA in the collection used. If
the collection used is, for instance, the TargetScan conserved
sites, there might nevertheless be non-conserved sites."),
tags$p("The plot size can be manually adapted by dragging the lower
right part with the mouse.")
),
tests=modalDialog(title="Enrichment tests", easyClose=TRUE,
.testIntro(), tags$p(
"For more information on the tests, please see the documentation."
)
),
tests2=modalDialog(title="Enrichment tests", easyClose=TRUE,
.testIntro(), .testDescription()
),
testsadvanced=modalDialog(title="Enrichment tests (advanced)", easyClose=TRUE,
.testDescription()),
noExample=modalDialog(easyClose=TRUE,
title="Example not available for this species", tags$p(
"There is unfortunately no example data available for this species.
You may however view an example by first switching to another species.")
),
background=modalDialog(easyClose=TRUE, title="Choice of the background",
tags$p("The background or universe defines the null hypothesis against
which the enrichment in the set of genes of interest is measured
and tested. As such, it applies only to tests that are based on
over-representation (e.g. siteoverlap), and not to tests that
are based on a continuous signal (e.g. areamir). In the context
of over-representation analysis, the background is critical to
meaningful results. An inappropriate background, in particular
a background that is too broad (e.g. all genes), will often
lead to spurious results."),
tags$p("The choice of the appropriate background depends on the
circumstances, but a key consideration is what genes ",tags$em(
"could have made it"), " into your selection. For example, if your
selection are genes that are differentially-expressed in a given
setting, then only genes that could have been differentially-
expressed should be included in the background. This means that
genes which are too lowly-expressed for any differential pattern to
be detected ought to be removed from the background.")
),
enrichplot_dl=modalDialog(easyClose=TRUE, title="Further customization",
tags$p("For further customization of the plot, simply ",
downloadLink("dl_hits2", "download the results table"),
" and load it in R using:", tags$br(),
tags$code('er <- read.csv("path/to/EnrichMir_hits_XXX.csv", row.names=1)')),
tags$p("You may then plot in full freedom with ", tags$code("ggplot2"),
"or use the enrichMiR wrapper. For the latter, first install ",
"it using:", tags$br(),
tags$code('BiocManager::install("ETHZ-INS/enrichMiR")'),tags$br(),
"(requires the 'remotes' package to be installed)"),
tags$p("and then use:", tags$br(), tags$code("library(enrichMiR)"),
tags$br(), tags$code("enrichPlot(er)"))
),
modalDialog(title=topic, "No help currently available for this topic.")
)
}
.testIntro <- function(){
tagList(
tags$p(
"enrichMiR implements different statistical tests for target
enrichment. Some tests depend on continous inputs (e.g. fold-changes),
and hence require the results of a differential expression analysis (DEA)
to be provided), while others are binary, i.e. based on the membership
in a gene set. When the input is a DEA, each binary tests will be
performed on both significantly upregulated and downregulated genes."
), tags$p(
"Note that the tests were developed and benchmarked for
application with predicted miRNA targets, not RNA binding proteins
(RBPs), which have more degenerate binding patterns. Although we
offer the possibility to perform RBP target enrichment analyses,
the tests were not benchmarked for that purpose, and the results
should therefore be interpreted with care."
), tags$p(
"Some tests gave poor performances in the benchmark, and should
therefore not be used. By default, the 'siteoverlap', 'woverlap' and
'areamir' tests are enabled (provided they are compatible with the
collection used), as these are the tests that gave the best performance.
When larger annotations are used (e.g. scanMiR or TargetScan-all), we
instead recommend using the 'lmadd' test, which showed the best performance
in this context."
)
)
}
.testDescription <- function(onlyDesc=FALSE){
desc <- tagList(
tags$p("The tests differ in the inputs they use, both in terms of the
target annotation as well as of the type of signal in which enrichment is
looked for. On the signal side, tests denoted as 'binary' compare features
(genes or transcripts) in a given set (e.g. your significantly downregulated
genes) to those in a background set (i.e. over-representation analysis),
whereas tests denoted as 'continuous' instead rely on a numeric input
signal, such as the mangitude or significance of changes in an input
differential expression analysis (by default, the tests use the sign of the
foldchange multiplied by the -log10(FDR), which is well correlated to logFC
for genes with low intra-group variability, and more robust than the latter).
On the annotation side, tests can also either use set membership (i.e.
whether or not a given feature is a predicted miRNA target) or numeric
values, such as the number of binding sites harbored by a given feature,
or a repression score (i.e. the extent to which a given feature is
predicted to be repressed by a miRNA)."),
tags$p("As is the case for over-representation analysis in general, for
tests based on binary inputs the choice of a good background is
critical. In many contexts, the background set of genes will be the
set of genes expressed in the system of interest."),
tags$ul(tags$li(tags$h4("Default tests"),
tags$ul(
tags$li(tags$b("siteoverlap")," (binary signal, set membership):",
tags$br(), "The siteoverlap test is based on Fisher's
exact test, but using the number of sites on predicted
targets and in the background instead of counting each
feature as one. While in theory this violates the
assumption of independence of the counts (since all the
binding sites of a given transcript are either in or out
of the set), leading to slightly anti-conservative
p-values, in practice this test is excellent at
identifying the most enriched miRNA."),
tags$li(tags$b("woverlap")," (binary signal, set membership):",
tags$br(), "This test is like the above 'siteoverlap' test,
but corrects for UTR length using the Wallenius method, as implemented in
the ", tags$a("goseq", target="_blank",
href="https://www.bioconductor.org/packages/devel/bioc/html/goseq.html"),
"package. The test performs similarly to the siteoverlap test.
Note that it requires a certain number of sets with overlaps to be run."),
tags$li(tags$b("areamir")," (continuous signal, score or set
membership):", tags$br(),
"The areamir test is based on the analytic Rank-based
Enrichment Analysis (aREA) test implemented in the
'msviper' function of the ", tags$a("viper", target="_blank",
href="https://www.bioconductor.org/packages/devel/bioc/html/viper.html"),
" package. The test is akin to an analytical version of
GSEA (see below), but it can additionally use degrees or
likelihood of set membership. If repression scores are
available in the annotation, areamir will therefore use a
(trimmed) version of it as set membership likelihood.")
)),
tags$li(tags$h5("Other tests implemented and evaluated"), tags$ul(
tags$li(tags$b("overlap")," (binary signal, set membership):",
tags$br(), "This test is based on Fisher's exact test, using the
number of features (i.e. transcripts/genes) among predicted targets
vs in the background (and therefore ignoring any site-based
information)."),
tags$li(tags$b("Mann-Whitney (MW)")," (continuous signal, set membership):",
tags$br(), "This is the Mann-Whitney (also known as Wilcoxon)
non-parametric test comparing targets and non-targets. This
test performs badly in benchmarks and should not be used."),
tags$li(tags$b("Kolmogorov-Smirnov (KS)")," (continuous signal, set membership):",
tags$br(), "This is the Kolmogorov-Smirnov test comparing the
signal distribution of targets vs non-targets. This
test performs badly in benchmarks and should not be used."),
tags$li(tags$b("modscore")," (continuous signal, repression score):",
tags$br(), "This is a linear regression testing the relationship
between the input signal and the corresponding repression score
predicted for a given miRNA."),
tags$li(tags$b("ebayes")," (continuous signal, repression score):",
tags$br(), "This is akin to the `modscore` tests, but performed
using limma's moderated t-statistics."),
tags$li(tags$b("lmadd")," (continuous signal, repression score):",
tags$br(), "This is the `ebayes` tests, followed by consecutive
fits adding each top miRNAs to the previous ones in a single
model. This is especially useful to identify candidates which
are not redundant with the top hit."),
tags$li(tags$b("modsites")," (continuous signal, number of sites):",
tags$br(), "This is a linear regression testing the relationship
between the input signal and the number of predicted binding
sites for a given miRNA, correcting for UTR length."),
tags$li(tags$b("GSEA"), " (continuous signal, set membership)",
tags$br(), "This test uses the multi-level fast GeneSet Enrichment
Analysis (GSEA) implemented in the ",tags$a("fgsea", target="_blank",
href="https://www.bioconductor.org/packages/devel/bioc/html/fgsea.html"),
"package, which is highly successful for Gene Ontology enrichment
analysis. In the context of our benchmark, however, it performed very
poorly."),
tags$li(tags$b("regmir"), tags$br(),
"The regmir test uses constrained lasso-regularized regression,
which has a high specificity but lower sensitivity. The test will use
binary or continuous inputs (using then either linear or binomial
regression), as well as binary set membership or predicted repression
score, depending on the availability of the input. The binary version
of the test has shown the best performances.")
))
)
)
if(onlyDesc) return(desc)
tagList(
tags$h3("Description of the enrichment tests"),
tags$p("Several target enrichment tests were benchmarked (results in the
benchmark tab), and are described below. The best overall tests were
selected as default for the app, and although other implemented tests are
made available in the app's advanced options, their usage is not
recommended."),
desc
)
}
.getTestsTable <- function(){
data.frame(
test=c("overlap","woverlap","siteoverlap","regmir (binary)","areamir","modSites",
"ebayes","lmadd","regmir (continuous)","KS","MW","GSEA"),
"input signal type"=rep(c("binary (set)", "continuous (DEA)"), c(4,8)),
"annotation type"=c("genesets","nb sites","nb sites","genesets","scores",
"nb sites","scores","scores","scores",rep("genesets",3)),
"description"=c(
"Over-representation (ORA) of target genes among set",
"ORA of binding sites, correcting for UTR length",
"ORA of binding sites",
"Regularized constrained logistic regression",
"Score-weighted analytic rank enrichment analysis",
"Linear regression of logFCs on nb of binding sites",
"Linear regression of logFCs on predicted repression scores",
"Consecutive additive linear regression of logFCs on predicted repression scores",
"Regularized constrained lienar regression of logFCs on predicted repression scores",
"Kolmogorov-Smirnov (KS) test on logFCs",
"Mann-Whitney / Wilcoxon test on logFCs",
"Gene set enrichment analysis (GSEA)"))
}
.getBrowserCompTable <- function(){
data.frame(
"OS"=c("Linux","MacOS","Windows"),
"Version" = c("Ubuntu","Big Sur","10"),
"Chrome" = c("96.0","96.0","89.0"),
"Firefox" = c("95.0","95.0","86.0"),
"Microsoft Edge" = c("n/a","n/a","89.0"),
"Safari" = c("n/a","15.0","n/a")
)
}
.exampleBackground <- function(){
c("PLEKHB2", "FBXO21", "PIGS", "SLC7A1", "YKT6", "RABL6", "SLC1A5",
"OCRL", "SH3RF1", "SLC9A1", "SLC7A5", "SRPRA", "G6PC3", "ARL2",
"CTDSPL", "CSRP1", "TBC1D22B", "ZER1", "SLC52A2", "GNPDA1", "PRDM4",
"VAMP3", "TMEM87A", "SEPTIN2", "SHCBP1", "RELL1", "PRTFDC1",
"CPEB1", "PHKA1", "RNF38", "TMED8", "CASP7", "PROSER1", "LMNB2",
"DSTYK", "TMEM216", "RAVER1", "PRUNE1", "ASF1B", "NCDN", "TGFBRAP1",
"CS", "MTHFD2", "SFT2D1", "ARHGAP1", "IQGAP1", "ATN1", "CTDNEP1",
"VPS37B", "PLCD3", "PKM", "POLR3D", "SLC25A6", "PRKRA", "HECTD3",
"SULF1", "SERINC5", "DYNC1LI2", "BCAT2", "VPS4A", "AL356776.1",
"DMAC1", "MROH8", "MSH2", "PANX2", "PAQR9", "SWI5", "DHX9P1",
"RAP1GDS1", "OLFM2", "PGF", "TBC1D12", "TNFSF12", "ANP32BP1",
"OSGIN2", "LMCD1", "CNOT9", "TTC6", "YTHDF2", "BOD1", "ZMYM3",
"USF1", "STK4", "PYCR3", "AC104083.1", "NSUN5P2", "PTAFR", "CFAP58",
"CDKN2AIP", "ARPC2", "HMGN2P46", "PJA1", "HSPA6", "AC079416.3",
"MYH9", "MRTFB", "SDK1", "HNRNPA1P69", "RAB10", "BUD13", "NDUFA2",
"STK31", "SEC11B", "ROMO1", "KDELC1P1", "EXOSC5", "AC092807.3",
"AC004492.1", "PRSS53", "MTRNR2L6", "RPS2P6", "RPL5P1", "AC013391.2",
"PNPT1P1", "CALM2", "AC027801.2", "ANO8", "RAB2B", "AL136038.4",
"PPM1B", "BHLHE41", "SBF1", "LIFR-AS1", "PSMB4", "POGK", "CSDC2",
"PPIAP26", "PHRF1", "NPM1P9", "PDIA2", "BCRP2", "GAPDHP58", "CLEC16A",
"AC234775.3", "EP400", "VTI1B", "PTPN6", "SIMC1", "AL034397.1",
"RBBP8", "PEX14", "ABCA11P", "TEN1-CDK3", "ZYG11B", "AGO1", "TAF3",
"GBX2", "PLIN5", "PKDCC", "SMNDC1", "RUSC2", "FAM20C", "HBP1",
"SDHAF1", "ZBTB40-IT1", "TMEM14C", "TUFM", "SH3GL2", "MDM4",
"MIR3936HG", "FAM210A", "SNRPGP10", "CCAR1", "RORB", "AL645940.1",
"RPN1", "LSM6", "PPRC1", "CD70", "AC027307.3", "ZNF789", "BRSK2",
"MEGF6", "DPYSL3", "CRYBG3", "DMAP1", "TMEM101", "CDKN1B", "UCP1",
"ZNF222", "LINC00926", "RGS7", "EML4", "PCBP1", "BTF3P8", "INTS1",
"AC106786.1", "RNF168", "CASP9", "MLXIPL", "CDX2", "ANKRD54",
"AC027682.7", "PRRG1", "AC209007.1", "AC026356.1", "AC104964.3",
"THG1L", "AC139530.1", "TEX22", "RPL7P59", "CRB2", "PDCD2", "HTR1D",
"PPIA", "YARS2", "AC092183.1", "TVP23C", "EEF1A1P6", "DNAAF1",
"INAFM1", "TRIOBP", "PIR", "NOP10", "DDX50P2", "ACTL8", "UQCRC2",
"RAB7A", "FBXL19-AS1", "PA2G4P4", "MAP4K2", "HADHAP2", "CKAP5",
"ZNF215", "GPX4", "MTX1", "TTLL4", "AHNAK", "KRT18P31", "ELP4",
"NCAM1", "DOK4", "ACAA2", "AC092431.2", "RPL22P8", "SP9", "ST6GALNAC6",
"TRA2A", "NOP2", "PLXNB2", "GADD45GIP1", "ASH1L-AS1", "ZNF843",
"RPS23P8", "KARS", "ZNF407", "TRAPPC2B", "ZNF518A", "KCNA3",
"UCKL1", "NPM1P6", "NOL3", "SPR", "NOP14", "LGMN", "VWA5A", "GAPDHP25",
"ARTN", "KBTBD7", "DCXR", "ANO6", "AC008147.4", "HAX1", "NEDD1",
"ZNF772", "AC020910.5", "DEPDC7", "HERC2P2", "DCHS1", "RPL37P1",
"PLIN1", "APLF", "ASB13", "NOP56", "PCSK4", "AMACR", "PLCL1",
"G0S2", "VASN", "PPP3CB", "PSMD4", "AL512791.2", "STK17A", "TSSK3",
"AC006460.2", "PPP1R26", "ELF1", "ACTBP8", "CSNK2B", "EDRF1",
"AC027097.1", "PTBP1", "KLLN", "GSTK1", "VKORC1L1", "COL11A2",
"C15orf62", "PRKCG", "ACVR2B-AS1", "RDH16", "WASHC2C", "AC114980.1",
"TXN2", "RPL19P16", "SRR", "RPS19BP1", "AC010680.5", "RN7SL832P",
"PCNA", "AC013403.2", "UBE2Q2", "AC016737.1", "TMPO", "API5P2",
"FAAH2", "STAMBP", "AL445487.1", "RPS2P44", "SRRM2-AS1", "AC079193.2",
"GATAD2A", "FEZF1", "HHEX", "RPF1", "IRX3", "AC093788.1", "STON2",
"TCN2", "CMSS1", "RPL27AP5", "DPH1", "AL136116.3", "TOMM5", "RPL14",
"AL365295.1", "AC090015.1", "VPS13B", "MXD4", "CHP2", "RN7SL535P",
"ACTG1P14", "AC087343.1", "PRPF38AP2", "GRIPAP1", "AC107068.1",
"CTSL", "CTSF", "SNHG11", "HHIP-AS1", "NMUR1", "R3HCC1L", "PPP1R13L",
"UGDH-AS1", "PDF", "GINS4", "PIPSL", "AC109347.1", "SLC16A10",
"RPS18P13", "ASPHD2", "EEF1E1P1", "AL096803.2", "RIC8B", "TFDP1",
"RBM6", "METTL4", "SULT4A1", "RPS2P46", "DSTN", "AC005072.1",
"ACAA1", "AC012313.1", "GSDME", "HNRNPA1P54", "TOGARAM2", "AC022498.2",
"TVP23B", "RPS12P26", "AC131235.1", "RCCD1", "INPP1", "WDFY1",
"AC026367.2", "MCF2L-AS1", "SETP21", "TMEM138", "SNORA33", "PABPN1",
"OGDH", "AL122023.1", "CHEK1", "CCN2", "AL139260.1", "SYMPK",
"CDKN2AIPNL", "MBD5", "NCOA4P4", "BLVRB", "DSE", "MIEF2", "RBM8B",
"ZNF350", "ADSL", "PARN", "LNCTAM34A", "PNMA6A", "PIGW", "TIMM29",
"IL21R-AS1", "NBPF3", "CYB5RL", "CERT1", "MAP7D2", "RUNX1", "AC127024.8",
"JAKMIP2", "NIBAN2", "TIGAR", "RNF227", "UNG", "MTOR", "TMEM107",
"JUND", "GEMIN7", "CALCOCO2", "TOR2A", "ATAD3A", "C1orf43", "CRK",
"AL353151.1", "GAPDHP23", "DEGS1", "KXD1", "AF131216.4", "GTF2H2C",
"LAMP2", "IPO7P1", "RAP2C-AS1", "LFNG", "MTND2P9", "CNOT7", "AC134349.1",
"GMPSP1", "NAA25", "MMP24OS", "VMP1", "NTNG2", "MCCC1", "RPS25",
"ILF2P2", "AL035530.2", "AC024451.1", "GNMT", "XRCC6P1", "NSMCE3",
"AC026124.1", "AC074194.1", "NHP2P1", "CHD2", "AC073046.1")
}
.exampleGeneset <- function(){
head(.exampleBackground(),60)
}
ETHZ-INS/enrichMiR documentation built on July 20, 2024, 12:03 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions .exampleGeneset .exampleBackground .getBrowserCompTable .getTestsTable .testDescription .testIntro .getHelpModal .getAppIntro
.getAppIntro <- function(){
data.frame(
element=c(NA, "#menu_species", "#collection_input", "#exprMirs_box",
"#expressed_mirna_outer", "#menu_input", "#input_type",
"#example_dea", "#menu_input", "#menu_enrich", "#enrichbtn_outer",
"#resultsbox", "#menu_cdplot", "#mir_fam", "#cdplot_outer", NA),
intro=c(
"This introduction will walk you through the usage of the enrichMiR app.
<br/><br/>
You can use the Next/Previous buttons to navigate the tutorial, and leave
it at any time by clicking outside it.",
"The first step is to make sure that you're working with the right species
and annotation. To set this, you open the 'Species and miRNAs' tab.",
"Once you've selected a species, you will be able to see and select one
of its available target annotations (see the help button for information
about the annotations). For the sake of this example, leave it at the
starting default, i.e. human 'targetScan conserved miRNA BS'.",
"This is optional, but you can also specify expressed miRNAs by expanding
this box (click the \"+\" on the right).<br/>This will be used to
restrict the miRNAs considered for enrichment analysis. In addition, you
will be able to visualize given expression values in the results.",
"There are three ways to provide this information:<br/>
1) you can specify a list of expressed miRNAs in 'Custom set';<br/>
2) you can upload a table of miRNA expression levels;<br/>
3) you can pick pre-compiled miRNA expression profiles for your
tissue/celltype of interest.<br/><br/>Not that you do not necessarily
need to specify expressed miRNA, but this is likely to give you better
results.",
"Once this is done, the next step is to provide the signal in which
enrichment should be looked for (e.g. your genes of interest, or
differential expression signature). To do this, click on the Input tab.",
"There are two main types of input:<br/>
1) the results of a differential expression analysis (DEA), which you can
upload using the 'Upload DEA results' tab, or<br/>
2) a set of genes to be tested for over-representation (the set can be
specified either manually or by selecting genes belonging to a given GO
category)<br/>
When specifying a set of genes, these will be tested for
over-representation among targets in comparison to a background set of
genes (also known as 'universe'). This is critical to over-representation
analysis! See the corresponding help button for more information about the
choice of a background.<br/><br/>
For the purpose of this example, select the 'Upload DEA results' tab.",
"For this tutorial, we'll use an in-built example DEA, which you can load
by clicking the 'Use example DEA' button.<br/><br/>Once you've done so,
you will see on the right that it successfully read the DEA file.",
"If you've provided a valid input, the input menu on the left will show
a blue badge instead of a red one.",
"We're now ready to launch the enrichment analysis!<br/><br/>First go to
the 'Enrichment analysis' tab",
"Then click the 'Run enrichMir!' button. The analysis will take a moment.",
"Once it's over, an enrichment plot will appear, displaying the
results. You can hover on the points for more information on the top hits.
(For more information on the enrichment plot and its interpretation, click
the help button above it).<br/><br/>
You can also visualize (and download) the results as a table in the
'Results Table' tab.",
"Independently of an enrichment analysis, when your input is a
differential expression analysis (DEA) result you also have the
possibility to create cumulative distribution (CD) plots for predicted
targets of any miRNA. To try this out, open the 'CD plot' tab.",
"By default, a CD plot is created for the first miRNA in the alphabetical
list. To select your miRNA of interest, click in the selector, erase the
current miRNA (using the backspace button), and start typing, for example
'133a'. The miRNA compatible with your selection will show up, and you can
select the one you are interested in. Select a miR-133a-3p, which is a
member of the family whose targets showed the strongest enrichment in
this DEA.",
"After a moment, the cumulative distribution plot will appear, showing
the foldchange distributions of different types of targets and
non-targets. In the presence of a real miRNA effect (that is gene expression
changes were primarily caused by a microRNA), we expect to see a
gradual dose-response pattern (targets with stronger binding sites also
have stronger log-foldchanges), as is the case in this example.<br/><br/>
Note that we could also have gotten to this CD plot by clicking on the
corresponding point in the enrichment plot of the previous tab!",
"This concludes our tutorial. While there are additionnal options here
and there, these are the app's main functionalities."
)
)
}
.getHelpModal <- function(topic){
switch(topic,
overview=modalDialog(title="enrichMiR workflow overview", easyClose=TRUE, size = "l",
tags$img(src="overview.png",height = 525, width = 750,
style="display: block; margin-left: auto; margin-right: auto;"),
tags$head(tags$style("#shiny-modal img { max-width: 100%; }")),
tags$p("For further information regarding the individual steps consult the
specifi help pages or conduct the interactive tour."),
tags$p("This website is free and open to all users and there is no login requirement.")
),
collections=modalDialog(title="Binding sites collections", easyClose=TRUE,
tags$p(
"Each collection represents a set of binding sites (validated or predicted) for
(a set of) miRNAs or families. Not all collections are available in all species.
Here is a brief description of each type of collection:"),
tags$ul(
tags$li(tags$b("TargetScan: "),
"miRNA binding site predictions from ",
tags$a( href="http://www.targetscan.org","TargetScan",
target="_blank"),
" (version 8). These are available in two flavors: using all
sites, or only sites conserved across vertebrates. In both cases,
only one UTR per gene (typically the longest) is considered."),
tags$li(tags$b("ScanMiR: "),
"miRNA binding site predictions using the ",
tags$a(href="https://www.bioconductor.org/packages/devel/bioc/html/scanMiR.html",
"scanMiR", target="_blank"),
"algorithm. ScanMiR employs a transcript-level affinity prediction method and is
compatible with both gene and transcript IDs. To ensure fast computing times, we restricted
the scanMiR annotation to 7mer and 8mer binding sites in the 3'UTR."),
tags$li(tags$b("miRTarBase: "),
"Experimentally-validated miRNA targets from the ",
tags$a(href="https://mirtarbase.cuhk.edu.cn/~miRTarBase/",
"miRTarBase", target="_blank"),
"database. While these targets are more trustworthy, a
sufficient number of targets is available for only a
proportion of the miRNAs, and some targets might have been
missed due to the cellular context of the experiments.
Because this collection does not include individual sites,
only the simplest over-representation test is available."),
tags$li(tags$b("oRNAment: "),
"RNA binding protein (RBPs) bindings from the ",
tags$a(href="http://rnabiology.ircm.qc.ca", "oRNAment",
target="_blank"), " database.", tags$br(),
"Note that enrichMiR was developed for miRNA bindings, and its
usage for RBPs is experimental.")
)
),
deaformat=modalDialog(title="Required format of the DEA table", easyClose=TRUE,
tags$p("The app should recognized the differential expression tables
generated from common RNAseq DEA package and saved as
character-delimited files, assuming that the gene names/IDs
are in the first column (which is normally the case).
In case of problem, try to format your table in this way:"),
tags$pre(
"Gene,logFC,PValue,FDR
CERS2,-1.54,1.37e-19,8.85e-16
NR6A1,-1.76,3.92e-19,1.77e-15
TAGLN2,-1.92,4.47e-17,1.08e-13
FTL,-1.69,1.68e-15,2.05e-12
TPM4,-1.41,1.83e-10,6.86e-08
..."),
tags$p("The character delimiter needs not be commas, and should be
automatically recognized. With most target annotations, both Ensembl
IDs and gene symbols will be recognized, and the scanMiR annotation
additionally supports transcript-level DEAs with Ensembl transcript IDs."),
tags$p("Because the accuracy of the fold-change estimates increases
with the genes' read count, we recommend filtering the DEA table
to include only highly expressed genes (e.g. the top 5000) before
upload.")
),
enrichplot=modalDialog(title="Enrichment plot", easyClose=TRUE, tags$p(
"The enrichment plot shows the significance of the enrichment on the
y-axis, and the magnitude of the enrichment on the x-axis. For basic
overlap/over-representation tests, the enrichment is the fold
increase over the overlap expected by chance. For continuous tests
(based on the results of a differential expression analysis), the
enrichment represents a magnitude of association with the
log-foldchanges of the putative targets."),
tags$p("In general, we recommend regarding the statistical significance
of individual posttranscriptional regulators obtained via the
different enrichment tests rather as rank instead of interpreting
them as finite results. Hence high-ranking miRNAs/families are
suggested to exert a functional role in the gene set at glance,
though certainly need to be further validated."),
tags$p("The size of the
points is proportional to the number of targets with binding sites for
the given miRNA/family. If miRNA expression data was given, the
points are colored by this expression."),
tags$p("Hovering on the points/families will provide extra
information about the family members and the test statistics."),
tags$p("In case a DEA has been provided as input, double-clicking on a
miRNA/family will automatically open the corresponding CD-plot."),
tags$p("The plot size can be manually adapted by dragging the lower right part
with the mouse.")
),
browsercompatibility=modalDialog(title="Browser compatibility", easyClose=TRUE,
fluidRow(column(8, align="center", tableOutput("browsercomp")))
),
cdplot=modalDialog(title="Cumulative distribution (CD) plots", easyClose=TRUE,
tags$p(
"Cumulative distribution plots are used to visualize differences in
log-foldchange distributions between groups of genes/transcripts.
Briefly, the genes in each sets are ranked according to their
foldchange (the x-axis), while the y-axis shows the proportion of
genes in the set that have a foldchange below or equal to that given
on x. In this way, small shifts in foldchanges can easily be
visualized."), tags$p(
"There are different ways of splitting the genes into groups, which
can be selected using the 'Split by' dropdown list. We recommend
splitting by best binding type if this information is available in
the collection you are using, or otherwise splitting by predicted
repression score (if available). This is the default (Automatic)
procedure."),
tags$p(renderPlot({
CDplot(list("no site"=rnorm(600,mean=0,sd=0.8),
"7mer"=c(rnorm(200,0),rnorm(200, mean=-0.5)),
"8mer"=c(rnorm(100,0),rnorm(200, mean=-1))), size=1.3) +
scale_color_manual(values=enrichMiR:::.siteTypeColors()[
c("no site", "7mer", "8mer")]) + xlab("logFC") + xlim(-3,2)
})),
tags$p("In the presence of a real miRNA effect, we expect to see a
gradual dose-response pattern, as shown above, meaning that
targets with stronger binding sites also have stronger
log-foldchanges (i.e. lower logFC, in the case of an increased
miRNA activity, as in this example)."),
tags$p("If your CD plot shows very angular lines, rather than
lines that approach curves, this means that your sets do not contain
enough elements. An example of this is given below:"),
tags$p(renderPlot({
CDplot(list("no site"=rnorm(600,mean=0,sd=0.8),
"7mer"=c(rnorm(8,0),rnorm(15, mean=-0.5)),
"8mer"=c(-2.8, -0.5, 0.5)), size=1.3) +
scale_color_manual(values=enrichMiR:::.siteTypeColors()[
c("no site", "7mer", "8mer")]) + xlab("logFC") + xlim(-3,2)
})),
tags$p("In such cases, if you are using score intervals, try reducing
the number of sets in the plot options. If you are using sites or
site types, this means that the miRNA of interest does not have
enough sites of this type in your dataset. This can especially
happen when using TargetScan ", tags$em("conserved"), "sites (or
miRTarBase sites), in which case you can try using all predicted
sites instead."),
tags$p("Because the accuracy of the fold-change estimates increases
with the genes' read count, we recommend using only highly
expressed genes to plot cummulative distributions of predicted
miRNA bindings, e.g. the top 5'000 - 7'000 expressed genes of a
dataset."),
tags$p("Note that 'no site' indicates genes which are not annotated
to have a binding site for that miRNA in the collection used. If
the collection used is, for instance, the TargetScan conserved
sites, there might nevertheless be non-conserved sites."),
tags$p("The plot size can be manually adapted by dragging the lower
right part with the mouse.")
),
tests=modalDialog(title="Enrichment tests", easyClose=TRUE,
.testIntro(), tags$p(
"For more information on the tests, please see the documentation."
)
),
tests2=modalDialog(title="Enrichment tests", easyClose=TRUE,
.testIntro(), .testDescription()
),
testsadvanced=modalDialog(title="Enrichment tests (advanced)", easyClose=TRUE,
.testDescription()),
noExample=modalDialog(easyClose=TRUE,
title="Example not available for this species", tags$p(
"There is unfortunately no example data available for this species.
You may however view an example by first switching to another species.")
),
background=modalDialog(easyClose=TRUE, title="Choice of the background",
tags$p("The background or universe defines the null hypothesis against
which the enrichment in the set of genes of interest is measured
and tested. As such, it applies only to tests that are based on
over-representation (e.g. siteoverlap), and not to tests that
are based on a continuous signal (e.g. areamir). In the context
of over-representation analysis, the background is critical to
meaningful results. An inappropriate background, in particular
a background that is too broad (e.g. all genes), will often
lead to spurious results."),
tags$p("The choice of the appropriate background depends on the
circumstances, but a key consideration is what genes ",tags$em(
"could have made it"), " into your selection. For example, if your
selection are genes that are differentially-expressed in a given
setting, then only genes that could have been differentially-
expressed should be included in the background. This means that
genes which are too lowly-expressed for any differential pattern to
be detected ought to be removed from the background.")
),
enrichplot_dl=modalDialog(easyClose=TRUE, title="Further customization",
tags$p("For further customization of the plot, simply ",
downloadLink("dl_hits2", "download the results table"),
" and load it in R using:", tags$br(),
tags$code('er <- read.csv("path/to/EnrichMir_hits_XXX.csv", row.names=1)')),
tags$p("You may then plot in full freedom with ", tags$code("ggplot2"),
"or use the enrichMiR wrapper. For the latter, first install ",
"it using:", tags$br(),
tags$code('BiocManager::install("ETHZ-INS/enrichMiR")'),tags$br(),
"(requires the 'remotes' package to be installed)"),
tags$p("and then use:", tags$br(), tags$code("library(enrichMiR)"),
tags$br(), tags$code("enrichPlot(er)"))
),
modalDialog(title=topic, "No help currently available for this topic.")
)
}
.testIntro <- function(){
tagList(
tags$p(
"enrichMiR implements different statistical tests for target
enrichment. Some tests depend on continous inputs (e.g. fold-changes),
and hence require the results of a differential expression analysis (DEA)
to be provided), while others are binary, i.e. based on the membership
in a gene set. When the input is a DEA, each binary tests will be
performed on both significantly upregulated and downregulated genes."
), tags$p(
"Note that the tests were developed and benchmarked for
application with predicted miRNA targets, not RNA binding proteins
(RBPs), which have more degenerate binding patterns. Although we
offer the possibility to perform RBP target enrichment analyses,
the tests were not benchmarked for that purpose, and the results
should therefore be interpreted with care."
), tags$p(
"Some tests gave poor performances in the benchmark, and should
therefore not be used. By default, the 'siteoverlap', 'woverlap' and
'areamir' tests are enabled (provided they are compatible with the
collection used), as these are the tests that gave the best performance.
When larger annotations are used (e.g. scanMiR or TargetScan-all), we
instead recommend using the 'lmadd' test, which showed the best performance
in this context."
)
)
}
.testDescription <- function(onlyDesc=FALSE){
desc <- tagList(
tags$p("The tests differ in the inputs they use, both in terms of the
target annotation as well as of the type of signal in which enrichment is
looked for. On the signal side, tests denoted as 'binary' compare features
(genes or transcripts) in a given set (e.g. your significantly downregulated
genes) to those in a background set (i.e. over-representation analysis),
whereas tests denoted as 'continuous' instead rely on a numeric input
signal, such as the mangitude or significance of changes in an input
differential expression analysis (by default, the tests use the sign of the
foldchange multiplied by the -log10(FDR), which is well correlated to logFC
for genes with low intra-group variability, and more robust than the latter).
On the annotation side, tests can also either use set membership (i.e.
whether or not a given feature is a predicted miRNA target) or numeric
values, such as the number of binding sites harbored by a given feature,
or a repression score (i.e. the extent to which a given feature is
predicted to be repressed by a miRNA)."),
tags$p("As is the case for over-representation analysis in general, for
tests based on binary inputs the choice of a good background is
critical. In many contexts, the background set of genes will be the
set of genes expressed in the system of interest."),
tags$ul(tags$li(tags$h4("Default tests"),
tags$ul(
tags$li(tags$b("siteoverlap")," (binary signal, set membership):",
tags$br(), "The siteoverlap test is based on Fisher's
exact test, but using the number of sites on predicted
targets and in the background instead of counting each
feature as one. While in theory this violates the
assumption of independence of the counts (since all the
binding sites of a given transcript are either in or out
of the set), leading to slightly anti-conservative
p-values, in practice this test is excellent at
identifying the most enriched miRNA."),
tags$li(tags$b("woverlap")," (binary signal, set membership):",
tags$br(), "This test is like the above 'siteoverlap' test,
but corrects for UTR length using the Wallenius method, as implemented in
the ", tags$a("goseq", target="_blank",
href="https://www.bioconductor.org/packages/devel/bioc/html/goseq.html"),
"package. The test performs similarly to the siteoverlap test.
Note that it requires a certain number of sets with overlaps to be run."),
tags$li(tags$b("areamir")," (continuous signal, score or set
membership):", tags$br(),
"The areamir test is based on the analytic Rank-based
Enrichment Analysis (aREA) test implemented in the
'msviper' function of the ", tags$a("viper", target="_blank",
href="https://www.bioconductor.org/packages/devel/bioc/html/viper.html"),
" package. The test is akin to an analytical version of
GSEA (see below), but it can additionally use degrees or
likelihood of set membership. If repression scores are
available in the annotation, areamir will therefore use a
(trimmed) version of it as set membership likelihood.")
)),
tags$li(tags$h5("Other tests implemented and evaluated"), tags$ul(
tags$li(tags$b("overlap")," (binary signal, set membership):",
tags$br(), "This test is based on Fisher's exact test, using the
number of features (i.e. transcripts/genes) among predicted targets
vs in the background (and therefore ignoring any site-based
information)."),
tags$li(tags$b("Mann-Whitney (MW)")," (continuous signal, set membership):",
tags$br(), "This is the Mann-Whitney (also known as Wilcoxon)
non-parametric test comparing targets and non-targets. This
test performs badly in benchmarks and should not be used."),
tags$li(tags$b("Kolmogorov-Smirnov (KS)")," (continuous signal, set membership):",
tags$br(), "This is the Kolmogorov-Smirnov test comparing the
signal distribution of targets vs non-targets. This
test performs badly in benchmarks and should not be used."),
tags$li(tags$b("modscore")," (continuous signal, repression score):",
tags$br(), "This is a linear regression testing the relationship
between the input signal and the corresponding repression score
predicted for a given miRNA."),
tags$li(tags$b("ebayes")," (continuous signal, repression score):",
tags$br(), "This is akin to the `modscore` tests, but performed
using limma's moderated t-statistics."),
tags$li(tags$b("lmadd")," (continuous signal, repression score):",
tags$br(), "This is the `ebayes` tests, followed by consecutive
fits adding each top miRNAs to the previous ones in a single
model. This is especially useful to identify candidates which
are not redundant with the top hit."),
tags$li(tags$b("modsites")," (continuous signal, number of sites):",
tags$br(), "This is a linear regression testing the relationship
between the input signal and the number of predicted binding
sites for a given miRNA, correcting for UTR length."),
tags$li(tags$b("GSEA"), " (continuous signal, set membership)",
tags$br(), "This test uses the multi-level fast GeneSet Enrichment
Analysis (GSEA) implemented in the ",tags$a("fgsea", target="_blank",
href="https://www.bioconductor.org/packages/devel/bioc/html/fgsea.html"),
"package, which is highly successful for Gene Ontology enrichment
analysis. In the context of our benchmark, however, it performed very
poorly."),
tags$li(tags$b("regmir"), tags$br(),
"The regmir test uses constrained lasso-regularized regression,
which has a high specificity but lower sensitivity. The test will use
binary or continuous inputs (using then either linear or binomial
regression), as well as binary set membership or predicted repression
score, depending on the availability of the input. The binary version
of the test has shown the best performances.")
))
)
)
if(onlyDesc) return(desc)
tagList(
tags$h3("Description of the enrichment tests"),
tags$p("Several target enrichment tests were benchmarked (results in the
benchmark tab), and are described below. The best overall tests were
selected as default for the app, and although other implemented tests are
made available in the app's advanced options, their usage is not
recommended."),
desc
)
}
.getTestsTable <- function(){
data.frame(
test=c("overlap","woverlap","siteoverlap","regmir (binary)","areamir","modSites",
"ebayes","lmadd","regmir (continuous)","KS","MW","GSEA"),
"input signal type"=rep(c("binary (set)", "continuous (DEA)"), c(4,8)),
"annotation type"=c("genesets","nb sites","nb sites","genesets","scores",
"nb sites","scores","scores","scores",rep("genesets",3)),
"description"=c(
"Over-representation (ORA) of target genes among set",
"ORA of binding sites, correcting for UTR length",
"ORA of binding sites",
"Regularized constrained logistic regression",
"Score-weighted analytic rank enrichment analysis",
"Linear regression of logFCs on nb of binding sites",
"Linear regression of logFCs on predicted repression scores",
"Consecutive additive linear regression of logFCs on predicted repression scores",
"Regularized constrained lienar regression of logFCs on predicted repression scores",
"Kolmogorov-Smirnov (KS) test on logFCs",
"Mann-Whitney / Wilcoxon test on logFCs",
"Gene set enrichment analysis (GSEA)"))
}
.getBrowserCompTable <- function(){
data.frame(
"OS"=c("Linux","MacOS","Windows"),
"Version" = c("Ubuntu","Big Sur","10"),
"Chrome" = c("96.0","96.0","89.0"),
"Firefox" = c("95.0","95.0","86.0"),
"Microsoft Edge" = c("n/a","n/a","89.0"),
"Safari" = c("n/a","15.0","n/a")
)
}
.exampleBackground <- function(){
c("PLEKHB2", "FBXO21", "PIGS", "SLC7A1", "YKT6", "RABL6", "SLC1A5",
"OCRL", "SH3RF1", "SLC9A1", "SLC7A5", "SRPRA", "G6PC3", "ARL2",
"CTDSPL", "CSRP1", "TBC1D22B", "ZER1", "SLC52A2", "GNPDA1", "PRDM4",
"VAMP3", "TMEM87A", "SEPTIN2", "SHCBP1", "RELL1", "PRTFDC1",
"CPEB1", "PHKA1", "RNF38", "TMED8", "CASP7", "PROSER1", "LMNB2",
"DSTYK", "TMEM216", "RAVER1", "PRUNE1", "ASF1B", "NCDN", "TGFBRAP1",
"CS", "MTHFD2", "SFT2D1", "ARHGAP1", "IQGAP1", "ATN1", "CTDNEP1",
"VPS37B", "PLCD3", "PKM", "POLR3D", "SLC25A6", "PRKRA", "HECTD3",
"SULF1", "SERINC5", "DYNC1LI2", "BCAT2", "VPS4A", "AL356776.1",
"DMAC1", "MROH8", "MSH2", "PANX2", "PAQR9", "SWI5", "DHX9P1",
"RAP1GDS1", "OLFM2", "PGF", "TBC1D12", "TNFSF12", "ANP32BP1",
"OSGIN2", "LMCD1", "CNOT9", "TTC6", "YTHDF2", "BOD1", "ZMYM3",
"USF1", "STK4", "PYCR3", "AC104083.1", "NSUN5P2", "PTAFR", "CFAP58",
"CDKN2AIP", "ARPC2", "HMGN2P46", "PJA1", "HSPA6", "AC079416.3",
"MYH9", "MRTFB", "SDK1", "HNRNPA1P69", "RAB10", "BUD13", "NDUFA2",
"STK31", "SEC11B", "ROMO1", "KDELC1P1", "EXOSC5", "AC092807.3",
"AC004492.1", "PRSS53", "MTRNR2L6", "RPS2P6", "RPL5P1", "AC013391.2",
"PNPT1P1", "CALM2", "AC027801.2", "ANO8", "RAB2B", "AL136038.4",
"PPM1B", "BHLHE41", "SBF1", "LIFR-AS1", "PSMB4", "POGK", "CSDC2",
"PPIAP26", "PHRF1", "NPM1P9", "PDIA2", "BCRP2", "GAPDHP58", "CLEC16A",
"AC234775.3", "EP400", "VTI1B", "PTPN6", "SIMC1", "AL034397.1",
"RBBP8", "PEX14", "ABCA11P", "TEN1-CDK3", "ZYG11B", "AGO1", "TAF3",
"GBX2", "PLIN5", "PKDCC", "SMNDC1", "RUSC2", "FAM20C", "HBP1",
"SDHAF1", "ZBTB40-IT1", "TMEM14C", "TUFM", "SH3GL2", "MDM4",
"MIR3936HG", "FAM210A", "SNRPGP10", "CCAR1", "RORB", "AL645940.1",
"RPN1", "LSM6", "PPRC1", "CD70", "AC027307.3", "ZNF789", "BRSK2",
"MEGF6", "DPYSL3", "CRYBG3", "DMAP1", "TMEM101", "CDKN1B", "UCP1",
"ZNF222", "LINC00926", "RGS7", "EML4", "PCBP1", "BTF3P8", "INTS1",
"AC106786.1", "RNF168", "CASP9", "MLXIPL", "CDX2", "ANKRD54",
"AC027682.7", "PRRG1", "AC209007.1", "AC026356.1", "AC104964.3",
"THG1L", "AC139530.1", "TEX22", "RPL7P59", "CRB2", "PDCD2", "HTR1D",
"PPIA", "YARS2", "AC092183.1", "TVP23C", "EEF1A1P6", "DNAAF1",
"INAFM1", "TRIOBP", "PIR", "NOP10", "DDX50P2", "ACTL8", "UQCRC2",
"RAB7A", "FBXL19-AS1", "PA2G4P4", "MAP4K2", "HADHAP2", "CKAP5",
"ZNF215", "GPX4", "MTX1", "TTLL4", "AHNAK", "KRT18P31", "ELP4",
"NCAM1", "DOK4", "ACAA2", "AC092431.2", "RPL22P8", "SP9", "ST6GALNAC6",
"TRA2A", "NOP2", "PLXNB2", "GADD45GIP1", "ASH1L-AS1", "ZNF843",
"RPS23P8", "KARS", "ZNF407", "TRAPPC2B", "ZNF518A", "KCNA3",
"UCKL1", "NPM1P6", "NOL3", "SPR", "NOP14", "LGMN", "VWA5A", "GAPDHP25",
"ARTN", "KBTBD7", "DCXR", "ANO6", "AC008147.4", "HAX1", "NEDD1",
"ZNF772", "AC020910.5", "DEPDC7", "HERC2P2", "DCHS1", "RPL37P1",
"PLIN1", "APLF", "ASB13", "NOP56", "PCSK4", "AMACR", "PLCL1",
"G0S2", "VASN", "PPP3CB", "PSMD4", "AL512791.2", "STK17A", "TSSK3",
"AC006460.2", "PPP1R26", "ELF1", "ACTBP8", "CSNK2B", "EDRF1",
"AC027097.1", "PTBP1", "KLLN", "GSTK1", "VKORC1L1", "COL11A2",
"C15orf62", "PRKCG", "ACVR2B-AS1", "RDH16", "WASHC2C", "AC114980.1",
"TXN2", "RPL19P16", "SRR", "RPS19BP1", "AC010680.5", "RN7SL832P",
"PCNA", "AC013403.2", "UBE2Q2", "AC016737.1", "TMPO", "API5P2",
"FAAH2", "STAMBP", "AL445487.1", "RPS2P44", "SRRM2-AS1", "AC079193.2",
"GATAD2A", "FEZF1", "HHEX", "RPF1", "IRX3", "AC093788.1", "STON2",
"TCN2", "CMSS1", "RPL27AP5", "DPH1", "AL136116.3", "TOMM5", "RPL14",
"AL365295.1", "AC090015.1", "VPS13B", "MXD4", "CHP2", "RN7SL535P",
"ACTG1P14", "AC087343.1", "PRPF38AP2", "GRIPAP1", "AC107068.1",
"CTSL", "CTSF", "SNHG11", "HHIP-AS1", "NMUR1", "R3HCC1L", "PPP1R13L",
"UGDH-AS1", "PDF", "GINS4", "PIPSL", "AC109347.1", "SLC16A10",
"RPS18P13", "ASPHD2", "EEF1E1P1", "AL096803.2", "RIC8B", "TFDP1",
"RBM6", "METTL4", "SULT4A1", "RPS2P46", "DSTN", "AC005072.1",
"ACAA1", "AC012313.1", "GSDME", "HNRNPA1P54", "TOGARAM2", "AC022498.2",
"TVP23B", "RPS12P26", "AC131235.1", "RCCD1", "INPP1", "WDFY1",
"AC026367.2", "MCF2L-AS1", "SETP21", "TMEM138", "SNORA33", "PABPN1",
"OGDH", "AL122023.1", "CHEK1", "CCN2", "AL139260.1", "SYMPK",
"CDKN2AIPNL", "MBD5", "NCOA4P4", "BLVRB", "DSE", "MIEF2", "RBM8B",
"ZNF350", "ADSL", "PARN", "LNCTAM34A", "PNMA6A", "PIGW", "TIMM29",
"IL21R-AS1", "NBPF3", "CYB5RL", "CERT1", "MAP7D2", "RUNX1", "AC127024.8",
"JAKMIP2", "NIBAN2", "TIGAR", "RNF227", "UNG", "MTOR", "TMEM107",
"JUND", "GEMIN7", "CALCOCO2", "TOR2A", "ATAD3A", "C1orf43", "CRK",
"AL353151.1", "GAPDHP23", "DEGS1", "KXD1", "AF131216.4", "GTF2H2C",
"LAMP2", "IPO7P1", "RAP2C-AS1", "LFNG", "MTND2P9", "CNOT7", "AC134349.1",
"GMPSP1", "NAA25", "MMP24OS", "VMP1", "NTNG2", "MCCC1", "RPS25",
"ILF2P2", "AL035530.2", "AC024451.1", "GNMT", "XRCC6P1", "NSMCE3",
"AC026124.1", "AC074194.1", "NHP2P1", "CHD2", "AC073046.1")
}
.exampleGeneset <- function(){
head(.exampleBackground(),60)
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.