03b614da15d6d8f07f0b974735e84d8fc42bf39c
bnguy
  Fri Apr 3 14:25:45 2020 -0700
Update on projects and tools page regarding NS forest and pyle

diff --git src/hg/cirm/gateway/htdocs/tools.html src/hg/cirm/gateway/htdocs/tools.html
index 9d3b395..ded488d 100755
--- src/hg/cirm/gateway/htdocs/tools.html
+++ src/hg/cirm/gateway/htdocs/tools.html
@@ -154,33 +154,32 @@
           SCIMITAR was created Pablo Cordero.
           </small>
         </div>
       </div>
     </div>
   </div>
   <br>
   <div class="row no-gutters">
     <div class="card-deck">
       <a name="tagStorm"></a>
       <div class="card">
         <div class="card-body text-center">
           <a href="https://cirm.ucsc.edu/images/tagStormImage.png" target="_blank">
           <img class="card-top-img" style="max-width: 300px"
           src="<!--#echo var='ROOT'-->images/tagStormImage.png">
-          <p class="lead"><span style="color: #1E90FF">Tag Storm:</span> Metadata made
-          simple</p>
           </a>
+          <p class="lead"><span style="color: #1E90FF">Tag Storm:</span> Metadata made simple</p>
           <ul class="list-group list-group-flush text-left">
             <li class="list-group-item">
             <p>
             The tag storm format offers an easy way to describe
             a hierarchical set of metadata for your data. The tag storm format is made up of
             a number of key/value pairs spread across stanzas at various indentation levels.
             Lower indentation levels inherit information from higher levels, so the most general
             metadata, e.g. lab or dataset names, are generally found at the highest level and
             more specifc metadata, e.g. sex or file name, are found at the lower level.</p>
             <p>
             The Tag Storm GitHub page contains more information about the format as well as a
             suite of tools to work with and manipulate tag storm files.
             </li>
             <li class="list-group-item">
             <a href="https://github.com/ucscGenomeBrowser/kent/tree/master/src/tagStorm"
@@ -357,48 +356,51 @@
             href="https://www.ndexbio.org/#/"          
             target="_blank">NDEx Website</a> <i class="fas fa-external-link-alt" aria-hidden="true"></i></li>
             <li class="list-group-item"> PMID: <a 
             href="https://www.ncbi.nlm.nih.gov/pubmed/26594663" target="_blank">26594663</a>; PMC: <a
             href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4649937/" target="_blank">PMC4649937</a></li>   
           </ul>    
         </div>
         <div class="card-footer">
           <small class="text-muted">
           NDEx Project was created by the NDEx Team at UC San Diego.
           </small>
         </div>
       </div>
     </div>
   </div>
+  <br>
   <div class="row no-gutters">
     <div class="card-deck">
       <div class="card">
         <div class="card-body text-center">
           <br>
           <a href="https://avatars1.githubusercontent.com/u/12550239?s=200&v=4">
           <img class="card-img-top"Style="max-width: 300px" 
           src="https://avatars1.githubusercontent.com/u/12550239?s=200&v=4"> 
           </a>      
           <p class="lead"><span style="color: #1E90FF">NS-Forest:</span>
-          Necessary and Sufficient Forest (NS-Forest) for Cell Type Marker Determination from cell type clusters</p>
+          Necessary and Sufficient Forest (NS-Forest) for Cell Type Marker Determination</p>
           <ul class="list-group list-group-flush text-left">
             <li class="list-group-item">
             <p>
             Necessary and Sufficient Forest is a method that takes cluster results from single cell/nuclei RNAseq experiments and generates lists of minimal markers needed to define each <q>cell type cluster</q>.
             The method begins by re-encoding the cluster labels into binary classifications, and Random Forest models are generated comparing each cluster versus all. The top fifteen genes are then reranked using a score measuring how binary they are, e.g., a gene with expression in the target cluster but no expression in the other clusters would have a high binary score. Expression cutoffs for the top six genes ranked by binary score are then determined by generating individual decision trees and extracting the decision path information. Then all permutations of the top six most binary genes are evaluated using f-beta score as an objective function (the beta value default set at 0.5, which weights the f-measure score more toward precision as opposed to recall)
             </p>
             </li>
             <li class="list-group-item"><a href="https://github.com/JCVenterInstitute/NSForest"
             target="_blank">NSForest GitHub</a> <i class="fab fa-github" aria-hidden="true"></i></li>
           </ul>    
         </div>
         <div class="card-footer">
           <small class="text-muted">
           NSForest was created by Brian Aevermann and Richard Scheuermann at JCVenterInstitute.
           </small>
         </div>
       </div>
+      <div class="col-md-6">
+      </div>
     </div>
   </div>
 </div>
 
 <!--#include virtual="${ROOT}inc/footer.html" -->