scanMiRApp: scanMiR shiny application

.getAppIntro <- function(){
  data.frame(
    element=c(NA, "#menuCollection", "#menuSearch", "#menuMiRNA", "#mirnainput",
              "#affinitybox", "#targetsbox", NA, "#subjectsbox", "#txbox",
              "#mirnasbox", "#scanBtn", "#hitsbox", NA),
    intro=c(
      "This introduction will walk you through the usage of the scanMiR app.",
      "First, use this tab to select the miRNA collection relative to your 
      species of interest",
      "In this section, you can scan and visualize specific transcripts or 
      sequence for miRNA binding sites",
      "In this section, you can instead start from a specific miRNA, visualize 
      its binding profile, and get its top predicted targets.<br/><br/>
      We'll start by looking at that part.",
      "Use this dropdown to select the miRNA you're interested in.<br/><br/>
      Note that you don't have to scroll down until you find it - you can 
      simply erase what's written in the box, type the beginning of the miRNA 
      name, and see the matching options.",
      "This box contains a plot summarizing the binding profile of the miRNA, 
      plotting the dissociation rate (i.e. affinity) of the top 7mers sequences
      (with or without the 'A' at position 1).<br/><br/>If it's not showing, 
      it's because the box is collapsed - you can open it by clicking the 'plus'
      button on the right.",
      "This box contains the predicted repression and number of binding sites 
      for all transcripts. You can reorder it anyway you like, and filter or 
      remove any column.<br/><br/>You can also double-click on one of the row 
      to get the details and visualize the individual binding sites - this would
      automatically move us to the transcript-centered section of the app.",
      "Moving on to the <b>transcript-centered</b> section!<br/><br/>
      The '<i>Search in gene/sequence</i>' menu on the left provides access to 
      sub-sections, the first of which ('subject') is about selecting the 
      sequence we want to scan.",
      "There are two ways to select a sequence to scan, available through the 
      tabs on the top right:<br>
      you can either scan an annotated transcript, or a custom sequence that 
      you'll copy-paste. We'll look at the former.",
      "In this sub-tab, you can first select the gene and then select the 
      transcript.<br/>Again, you can delete the content of the box and type the 
      first few letters to get the matching options.<br/>
      An overview of the selected sequence is then shown at the bottom.",
      "The next step is to select the miRNA(s) for which you want to scan 
      binding sites.<br/><br/>
      You can click in the box again and type the beginning of the miRNA(s) you
      wish to add and the matching miRNAs will show up. Alternatively, you can 
      select a set of miRNAs using the buttons.",
      "Finally, you can click on the scan button to launch it!<br/>
      If it's disabled, it's most likely because you didn't select any miRNA 
      and/or sequence.",
      "The putative binding sites are shown in the <i>'hits'</i> tab.<br/>
      You've got two ways to browse sites: you can either visualize them on a 
      plot along the length of the sequence, or through the table.<br/>
      In both cases, you can <i>double click</i> on a given match to view the 
      supplementary 3' alignment on the target sequence.",
      "Well, that's about it for the basic functions! Please let us know if you
      encounter problems!"
    )
  )
}

.getHelpModal <- function(topic){
  switch(topic,
    hitsCol=showModal(modalDialog(title="Columns of the hits table:",
      tags$p("The start and end columns represent the coordinates of the seed
binding site (i.e. corresponding to positions 1-8 of the miRNA) on the
target sequence. The coordinates are based on the beginning of the sequence
scanned (i.e. if you scanned only the UTR region, 1 represents the first
nucleotide of the UTR)."), tags$p("
The 'type' column indicates the type of match in the seed region."), tags$p("
The 'log_kd' column indicates the log of the estimated dissociation rate. A
lower log_kd value is indicative of a stronger affinity."), tags$p("
The 'p3.score' represent the strength of the 3' supplementary alignment. In
most cases, it roughly corresponds to the number of consecutive complementary
bases (exluding the seed region). However, because of tolerance for G-U
bindings or mismatches in large regions of complementarity, it can depart from
this rule of thumb. To visualize this alignment, you can simply double click
on one of the rows of the table."), tags$p("
The 'note' column contains eventual special features of the binding site (e.g.
prediction of TDMD sites)."), easyClose=TRUE, footer=NULL)),
    stypes=showModal(modalDialog(easyClose=TRUE, footer=NULL, size="l",
      title="Description of the site types",
      tags$p("miRNAs can bind to mRNAs in different kinds, generally defined by
        the 'binding site types'. See the figure below from ", 
      tags$a(href="https://doi.org/10.1016/j.cell.2009.01.002", 
             "Bartel (2009)", target="_blank"), "for a first overview and 
        the corresponding  review for detailed information."),
      tags$div(style="height: 576px;", imageOutput("bartel2009")), 
      tags$p("In brief, there are canonical and non-canonical binding sites. The
        canonical binding sites are primarily classified by the number of 
        consecutive nucleotides of the extended seed region of a miRNA that are 
        bound to their complements. '7mer-m8' indicates that the 8th nucleotide 
        (nt) of the miRNA is bound at that binding site, '7mer-A1' that there 
        is an 'A' opposite the first nt of the miRNA, which is beneficial at 
        that specific position. Non-canoncial binding sites typically include 
        parts of the seed, sometimes with wobble bindings or bulged out 
        nucleotides, but often do not cause effective silencing of transcripts.
        An overview of the affinity of the top sites can be seen in the miRNA
        tab."),
      tags$p(tags$a(href="https://dx.doi.org/10.1126/science.aav1741", 
        "McGeary, Lin et al. (2019)", target="_blank"), " show that in order to
        determine the efficacy of a binding site, it's important to not only 
        consider the type of the site but also take the flanking di-nucleotides
        into account. Finally, in some cases supplementary binding on the miRNA
        3' side can increase the stability of the binding or lead to other
        behaviors (such as target-mediated miRNA degradataion), however the 
        impact of such supplementary binding is difficult to predict.")
    )),
    manhattan=showModal(modalDialog(easyClose=TRUE, footer=NULL,
      title="Plotting sites along transcript",
      tags$p("This plot shows the predicted miRNA binding sites at their 
        positions (on the x-axis) within the selected sequence. The '-log_kd' 
        (= -log(KD)) information on the y-axis is a proxy for the  affinity of 
        a specific miRNA to its complementary sequence at that position. The 
        higher a '-log_kd' value is, the more time the Ago-miRNA complex will 
        spend at that site and the more likely this particular binding site 
        will lead to post-transcriptional repression of the transcript."),
      tags$p("If the scan has been performed with a single miRNA, a green area 
        highlighting will show the affinity range of the strongest site types 
        for this miRNA (i.e. the -log_kd range of 8mers), which can help assess
        the strength of individual binding sites."),
      tags$p("If the scan has been performed with several miRNAs, by default 
        only those with the strongest sites (see also the order in the 
        hits-table) will be displayed in the plot. The number of shown miRNAs 
        can be altered in the 'Number of top miRNAs displayed' input"),
      tags$p("A 'log_kd' of -1 (indicated by the dashed red line) indicates 
             background binding."),
      tags$p("Note log_kd values are not necessarily comparable across miRNAs!")
    )),
    collections=showModal(modalDialog(easyClose=TRUE, footer=NULL,
      title="miRNA Kd-Model collections",
      tags$p("In this section you can select a miRNA species collection, which 
        will determine the available miRNA-KdModels that you can subsequently 
        choose to scan sequences or display in the miRNA-based tab. We 
        pre-compiled KdModels for all human, mouse and rat",
        tags$a(href="https://www.mirbase.org/index.shtml", "miRBase v22", 
          target="_blank"), "miRNAs. To get an overview of the number of 
        available miRNAs for the selected species and their",
        tags$a(href="http://www.targetscan.org/vert_80/", "Targetscan", 
          target="_blank"),"conservation status simply click on 'Details'."),
      tags$p("You can also run the App with your own list of KdModels if your 
        species of interest is for example not provided by default. Consult the
        ", tags$a(href="http://bioconductor.org/packages/release/bioc/html/scanMiRApp.html", 
        "vignette", target="_blank"),"of the Bioconductor package for further 
        information.")
    )),
    modalDialog(title=topic, "No help currently available for this topic.")
  )
}