graphclust_step_by_step.yaml

name: GraphClust workflow step by step
description: Step by step instructions for using GraphClust for clustering RNA sequences
title_default: "<b>GraphClust step by step</b>"
tags:
  - "GraphClust"

steps:
    - title: "<b>A tutorial on GraphClust(Clustering RNA sequences)</b>"
      content: "This tour will walk you through the process of <b>GraphClust</b> to cluster RNA sequences.<br><br>
                Read and Follow the instructions before clicking <b>'Next'</b>.<br><br>
                Click <b>'Prev'</b> in case you missed out on any step."
      backdrop: true

    - title: "<b>A tutorial on GraphClust</b>"
      content: "Together we will go through the following 10 steps:<br>
                 <dir>
                 <b>
                   <li>Data Acquisition</li>
                   <li>Pre-processing</li>
                   <li>Creating a graph from FASTA</li>
                   <li>Creating a sparse vector</li>
                   <li>Computing candidate clusters</li>
                   <li>Preprocessing data for computing best subtree</li>
                   <li>Computing best subtree using LocaRna</li>
                   <li>Finding consensus motives</li>
                   <li>Building Covariance Model using cmbuild</li>
                   <li>Searching homologous sequences using cmsearch</li>
                   <li>Post-processing</li>
                 </b>
                 </dir>"
      backdrop: true


    - title: "<b>Data Acquisition</b>"
      content: "We will start with a simple small <b>FASTA</b> file.<br><br>
                You will get one FASTA file with RNA sequences that we want to cluster.<br><br>"
      backdrop: true

    - title: "<b>Data Acquisition</b>"
      element: ".upload-button"
      intro: "We will import the FASTA file into into the history we just created.<br><br>
              Click <b>'Next'</b> and the tour will take you to the Upload screen."
      position: "right"
      postclick:
        - ".upload-button"

    - title: "<b>Data Acquisition</b>"
      element: "button#btn-new"
      intro: "The sample training data available on github is a good place to start.<br><br>
              Simply click <b>'Next'</b> and the links to the training data will be automatically inserted and ready for upload.<br><br>
              Later on, when you want to upload other data, you can do so by clicking the <b>'Paste/Fetch Data'</b> button or
              <b>'Choose local file'</b>  to upload locally stored file."
      position: "top"
      postclick:
        - "button#btn-new"

    - title: "<b>Data Acquisition</b>"
      element: ".upload-text-content:first"
      intro: "Links Acquired !"
      position: "top"
      textinsert:
        https://github.com/BackofenLab/docker-galaxy-graphclust/raw/master/data/Rfam-cliques-dataset/cliques-low-representatives.fa

    - title: "<b>Data Acquisition</b>"
      element: "button#btn-start"
      intro: "Click on <b>'Start'</b> to upload the data into your Galaxy history."
      position: "top"

    - title: "<b>Data Acquisition</b>"
      element: "button#btn-close"
      intro: "The upload may take awhile.<br><br>
              Hit the <b>close</b> button when you see that the files are uploaded into your history."
      position: "top"

    - title: "<b>Data Acquisition</b>"
      element: "#current-history-panel > div.controls"
      intro: "You've acquired your data. Now let's start using GraphCLust tools.<br><br>"
      position: "left"

    - title: "<b>GraphCLust</b>"
      intro: "Once we have the data for analysis we can start the process of clustering RNA sequences step by step. <br>
              Navigate to the tool panel on the left side and click on <b>GraphCLust</b>. This will open a section
              with set of tool necessary for GraphCLust process. <br>
              The first step of the process is <b>Preprocessing</b>."
      position: "right"


    - title: "<b>Pre-processing</b>"
      intro: "This tool takes as an input file of sequences in Fasta format
                and creates the final input for GraphCLust based on given parameters. Parameters allows us to
                split long sequences into smaller fragments to enable the detection of local signals"
      position: "right"
      postclick:
        - "#preproc > div.toolTitle > div > a"


    - title: "<b>Pre-processing</b>"
      element: "#s2id_uid-18_select > a"
      intro: "Here we should define our input file.<br>"
      position: "top"


    - title: "<b>Pre-processing</b>"
      element: "#uid-11"
      intro: "<b>'Window size'</b> defines the length of the fragments that the input sequences will be split.
              In default settings we set it to very high number to not split the sequences at all.<br><br>"
      position: "top"


    - title: "<b>Pre-processing</b>"
      element: "#uid-64 > div.ui-form-title > span"
      intro: "<b>'Window shift in percent'</b> defines the percentage of the shift fot fragments of the input sequences.
              In default settings it is 100% because we don't split input sequences.<br><br>"
      position: "top"

    - title: "<b>Pre-processing</b>"
      element: "#uid-46 > div.ui-form-field"
      intro: "<b>Minimum sequence length</b> defines the minimal length of input sequences.<br><br>"
      position: "top"


    - title: "<b>Pre-processing</b>"
      element: "#execute"
      intro: "To run the tool press <b>'Execute' </b> button.<br>"
      position: "top"


    - title: "<b>Understanding the Output</b>"
      element: "#current-history-panel"
      intro: "After the tool is executed  several output files are created.
              <br> By clicking on the <b>'eye'</b> icon you can see the content of the files.
              <br><br>"
      position: "left"

    - title: "<b>Preprocessing : DONE</b>"
      intro: "Once preprocessing of the data is done we can move on to the next step :create a graph from FASTA file.<br>"
      position: "right"

    - title: "<b>Creating a graph from FASTA</b>"
      intro: "To create a graph we need to use a tool called <b>fasta_to_gspan</b> which you
              can find in GraphClust section of tool panel.<br>"
      position: "right"
      postclick:
        - "#gspan > div.toolTitle > div > a"

    - title: "<b>Creating a graph from FASTA</b>"
      element: "#uid-71 > div.ui-form-field > div.ui-select-content > div.ui-options > div.btn-group.ui-radiobutton"
      intro: "As an input this tool takes pre-processed FASTA file: <b>'data.fasta'</b>.<br>"
      position: "right"


    - title: "<b>Creating a graph from FASTA</b>"
      intro: "Detailed description of each parameter you can find in help section at the bottom of the page"
      position: "top"

    - title: "<b>Creating a graph from FASTA</b>"
      element: "#execute"
      intro:  "To run the tool press <b>'Execute' </b> button.<br>"
      position: "top"

    - title: "<b>Creating a graph from FASTA</b>"
      intro:  "Once tool is executed it will produce <b>gspan.zip file</b>.<br>"
      position: "top"

    - title: "<b>Creating a graph from FASTA : DONE</b>"
      intro:  "Now we have <b>gspan.zip</b> file and can move on the next step.<br>
               The next step is to create sparse vector using NSPDK. <br>
               From GraphClust tools choose <b>NSPDK_sparseVect</b> to start the process. <br>"
      position: "top"
      postclick:
        - "#nspdk_sparse > div.toolTitle > div > a"

    - title: "<b>Creating a sparse vector</b>"
      intro:  "This tool will create explicit sparse feature encoding using <b>NSPDK</b>.<br>"
      position: "top"


    - title: "<b>Creating a sparse vector</b>"
      intro:  "This tools requires 2 input files:
                <br>
                   <dir>
                   <b>
                     <li>data.fasta : from pre-processing step</li>
                     <li>gspan.zip : from the previous step</li>
                   </b>
                   </dir>"
      position: "top"


    - title: "<b>Creating a sparse vector</b>"
      intro:  "More information about NSPDK you can find in help section<br>"
      position: "top"

    - title: "<b>Creating a sparse vector</b>"
      element: "#execute"
      intro:  "Run the tool by pressing <b>'Execute' </b> button.<br>"
      position: "top"

    - title: "<b>Creating a sparse vector</b>"
      intro:  "After tool is exec it will produce <b>data_svector</b> which we will use in the next step."
      position: "top"

    - title: "<b>Creating a sparse vector: DONE</b>"
      intro:  "Created <b>data_svector</b> file will be used in the next step for computing
              candidate cluster. For that click on <b>NSPDK_candidateClusters</b> tool."
      position: "top"
      postclick:
        - "#NSPDK_candidateClust > div.toolTitle > div > a"


    - title: "<b>Computing candidate clusters</b>"
      intro:  "During this step we will compute global feature index and get top dense sets.
              The candidate clusters are chosen as the top ranking neighborhoods provided that the
                size of their overlap is below a specified threshold."
      position: "top"



    - title: "<b>Computing candidate clusters</b>"
      intro:  "Here we have to specify 3 input files:
                <br>
                   <dir>
                   <b>
                     <li>data_svector : from the previous step</li>
                     <li>data_fasta </li>
                     and
                     <li>data_names from pre-processing step</li>
                   </b>
                   </dir>"
      position: "top"

    - title: "<b>Computing candidate clusters</b>"
      intro:  "Another important parameter for this tool is <b>Multiple iterations</b>.
              This parameter by default is set to <b>'no'</b> which means we will do only a single iteration.
              By setting it to <b>'yes'</b> we have to define some other input files which we would get from previous
              iterations. But in the scope of this tutorial we will just go for a single iteration."
      position: "top"

    - title: "<b>Computing candidate clusters</b>"
      intro:  "For more information about this tool check the <b>help</b> section.<br>"
      position: "top"

    - title: "<b>Computing candidate clusters</b>"
      element: "#execute"
      intro:  "Run the tool by pressing <b>'Execute' </b> button.<br>"
      position: "top"

    - title: "<b>Computing candidate clusters</b>"
      intro:  "This step will produce 3 output files:
          <br>
             <dir>
             <b>
               <li>fast_cluster</li>
               <li>fast_cluster_sim</li>
               <li>blacklist</li>
             </b>
             </dir>
             <b>fast_cluster</b> represents the id's of the candidate clusters. <br>
             <b>fast_cluster_sim</b> represents the similarity scores. <br>
             <b>blacklist</b> contains the ids of sequences that were already clustered, that's why in case
             of the single iteration it's empty. <br>"
      position: "top"

    - title: "<b>Computing candidate clusters : DONE</b>"
      intro:  "Once output files are ready we can move to the nexr step by clicking on
                <b>premlocarna</b> tool.<br>"
      position: "top"
      postclick:
        - "#preMloc > div.toolTitle > div > a"

    - title: "<b>Preprocessing data for computing best subtree</b>"
      intro:  "This tool will do some pre-processing for computing best subtrees.<br>"
      position: "top"


    - title: "<b>Preprocessing data for computing best subtree</b>"
      intro:  "This step needs 4 input files:
                <br>
                   <dir>
                   <b>
                     <li>fast_cluster : from the previous step</li>
                     <li>fast_cluster_sim : from the previous step</li>
                     <li>data_fasta </li>
                     and
                     <li>data_names from pre-processing step</li>
                   </b>
                   </dir>
                   "
      position: "top"

    - title: "<b>Preprocessing data for computing best subtree</b>"
      element: "#execute"
      intro:  "Run the tool by pressing <b>'Execute' </b> button.<br>"
      position: "top"

    - title: "<b>Preprocessing data for computing best subtree</b>"
      intro:  "Execution of this tool results in 4 datasets :
          <br>
             <dir>
             <b>
               <li>centers</li>
               <li>trees</li>
               <li>tree_matrix</li>
               <li>cmfinder_fa</li>
               <li>model_tree_fa</li>
             </b>
             </dir>
            These datasets will be used in next steps.
        <br>"
      position: "top"

    - title: "<b>Preprocessing data for computing best subtree : DONE</b>"
      intro:  "Preprocessing for the best tree computation is done, so we can now do the actual computation
                of the best subtree. For that we need the tool called <b>locarna_best_subtree</b>.<br>"
      position: "top"
      postclick:
        - "#locarna_best_subtree > div.toolTitle > div > a"

    - title: "<b>Computing best subtree using LocaRna</b>"
      intro:  "This step computes a multiple sequence-structure alignment of RNA sequences using <b>LocaRna</b>.
              It uses tree file (from previous step) with guide tree in NEWICK format.
              The given tree is used as guide tree for the progressive alignment. It saves the calculation
              of pairwise all-vs-all similarities and construction of the guide tree. And at the end return the best subtree<br>"
      position: "top"


    - title: "<b>Computing best subtree using LocaRna</b>"
      intro:  "This step takes as an input following files:
         <dir>
         <b>
           <li>centers</li>
           <li>trees</li>
           <li>tree_matrix</li>
           from previous steps and
           <li>data_map from pre step</li>
         </b>
         </dir>
        <br>"
      position: "top"

    - title: "<b>Computing best subtree using LocaRna</b>"
      element: "#execute"
      intro:  "Run the tool by pressing <b>'Execute' </b> button.<br>"
      position: "top"

    - title: "<b>Computing best subtree using LocaRna</b>"
      intro:  "Output of this tool is <b>model.tree.stk</b> which will
                be used in next step to find consensus motives.<br>"
      position: "top"

    - title: "<b>Computing best subtree using LocaRna : DONE</b>"
      intro:  "Now we have <b>model.tree.stk</b> so we can find consensus motives using the next tool -
                <b>CMFinder_v0</b>.<br>"
      position: "top"
      postclick:
        - "#cmFinder > div.toolTitle > div > a"

    - title: "<b>Finding consensus motives</b>"
      intro:  "During this step conversion from CLUSTAL format files to STOCKHOLM format is done.
              Then using CMFinder we determine consensus motives for sequences.<br>"
      position: "top"


    - title: "<b>Finding consensus motives</b>"
      intro:  "This tool takes as an input the following files:
              <dir>
              <b>
                <li>model_tree_stk : from previous step</li>
                <li>cmfinder_fa : from 'Preprocessing data for computing best subtree' step </li>
                <li>tree_matrix</li>
                </b>
              </dir>
        <br>"
      position: "top"

    - title: "<b>Finding consensus motives</b>"
      element: "#execute"
      intro:  "Run the tool by pressing <b>'Execute' </b> button.<br>"
      position: "top"

    - title: "<b>Finding consensus motives</b>"
      intro:  "Output of the tool is in STOCKHOLM format and contains the consensus structure. <br>"
      position: "top"

    - title: "<b>Finding consensus motives : DONE</b>"
      intro:  "Once we have consensus structure we can build covariance model with <b>cmbuild</b> tool.
              For that click on the tool named <b>Build covariance models </b><br>"
      position: "top"
      postclick:
        - "#infernal > div:nth-child(3) > div > a "

    - title: "<b>Building Covariance Model using cmbuild</b>"
      intro:  "In this step we cm build a covariance model of an RNA multiple alignment.
              <b>cmbuild</b> uses the consensus structure to determine the architecture of the covariance model. <br>"
      position: "top"


    - title: "<b>Building Covariance Model using cmbuild</b>"
      intro:  "As an input for this tool we give 'model_cmfinder_stk' file containing consensus
                structure from the pre step. <br>
                For more information about this tool read the <b>help</b> section of the page. <br>"
      position: "top"

    - title: "<b>Building Covariance Model using cmbuild</b>"
      element: "#execute"
      intro:  "Run the tool by pressing <b>'Execute' </b> button.<br>"
      position: "top"

    - title: "<b>Building Covariance Model using cmbuild : DONE</b>"
      intro:  "After covariance model is built we can move on to <cmsearch>.
              Simply click on <b>Search covariance model(s)</b>.<br>"
      position: "top"
      postclick:
        - "#infernal > div:nth-child(4) > div > a"

    - title: "<b>Searching homologous sequences using cmsearch</b>"
      intro:  "<b>cmsearch</b> allows you to make consensus RNA secondary structure profiles,
              and use them to search nucleic acid sequence databases for homologous RNAs. <br>"
      position: "top"

    - title: "<b>Searching homologous sequences using cmsearch</b>"
      intro:  "As a <b>sequence database </b> we choose <b>data_fasta_scan</b> file generated during
                pre-processing. <br>
                Then to use <b>covariance model</b> generated in previous step you should
                select from <b>Subject covariance models</b> dropdown menu <b>Covariance model from your history</b> option."
      position: "top"

    - title: "<b>Searching homologous sequences using cmsearch</b>"
      element: "#execute"
      intro:  "Run the tool by pressing <b>'Execute' </b> button.<br>"
      position: "top"

    - title: "<b>Searching homologous sequences using cmsearch : DONE</b>"
      intro:  "Finally we reached the last step of our workflow!
                Click on <b>Report_Results</b> to do the final step. <br>"
      position: "top"
      postclick:
        - "#glob_report > div.toolTitle > div > a"

    - title: "<b>Post-processing</b>"
      intro:  "Final step of our workflow is <b>Post-processing</b>.<br>
              In this step we will report clusters and merge them if needed."
      position: "top"


    - title: "<b>Post-processing</b>"
      intro:  "This tool takes as an input the following files:
              <dir>
              <b>
                <li>FASTA.zip : from Preprocessing step</li>
                <li>cmsearch_results : from previous step </li>
                and
                <li>model_tree_files : from <b>'Preprocessing data for computing best subtree'</b> step</li>
              </b>
              </dir>
        <br>"
      position: "top"

    - title: "<b>Post-processing</b>"
      intro:  "The final output: <br>
              <b>'cluster.final.stat'</b> file contains general information about clusters,
                      e.g. number of clusters, number of sequences in each cluster etc.
                      <br> By clicking on the <b>'eye'</b> icon you can see the content of the file.
                    <br>
              <b>'CLUSTERS'</b> dataset collection contains one file for each cluster. <br>
              Each file contains information about sequences in that cluster. Each line in the file contains: cluster number,
              cm_score, sequence origin (whether it comes from model or from Infernal search) and sequence id.
              <br>"
      position: "top"

    - title: "<b>A tutorial on GraphClust workflow</b>"
      intro: "Thank You for going through our tutorial."
      backdrop: true