35ad0a3238e9a98b1b2872e0de88e87b0f519e3a lrnassar Fri Jan 13 17:05:29 2023 -0800 Clarifying that the trix searching is note case sensitive, and fixing an example that is misleading. Refs #30451 diff --git src/hg/htdocs/goldenPath/help/trix.html src/hg/htdocs/goldenPath/help/trix.html index e4c1096..bac4f7d 100755 --- src/hg/htdocs/goldenPath/help/trix.html +++ src/hg/htdocs/goldenPath/help/trix.html @@ -1,60 +1,61 @@ <!DOCTYPE html> <!--#set var="TITLE" value="Genome Browser Trix Indices" --> <!--#set var="ROOT" value="../.." --> <!-- Relative paths to support mirror sites with non-standard GB docs install --> <!--#include virtual="$ROOT/inc/gbPageStart.html" --> <h1>Trix Indices</h1> <p> A Trix index consists of a pair of files that allow for fast look-up of free text associated with a list of identifiers. The index is created from a single line-oriented text file using the program <code>ixIxx</code>. Each line in the text file starts with an identifier, followed by free text -associated with the ID. For a more complete description of how to make a searchable track hub (or +associated with the ID. The search is not case sensitive, so any case-combination of the free text +entered will me matched. For a more complete description of how to make a searchable track hub (or custom track), please visit the <a href="hubQuickStartSearch.html" target="_blank">Quick Start Guide to Searchable track hubs</a>.</p> <p> To complete the steps below you must first download the <code>ixIxx</code> utility. For more information on downloading our command line utilities, see these <a href="http://hgdownload.soe.ucsc.edu/downloads.html#source_downloads">instructions</a>.</p> <h3 id="ex1">Example 1</h3> <p> To create a Trix index, follow these steps: <ol> <li> Prepare a text file that associates your IDs with free text:<br> <pre><code> id1 this is text for id1 id2 this is text for id2 id3 this is text for id3</code></pre></li> <li> Run the <code>ixIxx</code> program on your text file.<br> <pre><code> ixIxx input.txt myTrix.ix myTrix.ixx</code></pre></li> </ol> <h3 id="ex2">Example 2</h3> <ol> <li> If you have a bigBed track with unique gene names such as <code>SIRT1, BRCA1, TP53</code> in the fourth name column, and you built the bigBed using the option of <code>-extraIndex=name</code> to index the name field you can create an input.txt such as the following that associates the bigBed name with other identifiers people might search for or with shorter spellings of the names: <pre> - SIRT1 sirt1 sir sirt SIR SIRT Sirtuin SIR2-Like ENSG00000096717 NM_012238 - BRCA1 brca1 brca brc BRC BRCA breast cancer 1 ENSG00000012048 NM_007300 - TP53 tp53 tp5 TP5 Tumor Protein P53 ENSG00000141510 NM_001126112</pre></li> + SIRT1 sirt1 sir sirt Sirtuin SIR2-Like ENSG00000096717 NM_012238 + BRCA1 brca1 brca brc breast cancer 1 ENSG00000012048 NM_007300 + TP53 tp53 tp5 Tumor Protein P53 ENSG00000141510 NM_001126112</pre></li> <li> You would then run the <code>ixIxx</code> program on your text file, taking into account the length of your longest word.<br> <pre><code> ixIxx input.txt myTrix.ix myTrix.ixx</code></pre></li> </ol> <p> The <code>ixIxx</code> utility has a default of 31 characters for words. If you build an input.txt with longer words, such as very long accessions, you neeed to add the option <code>-maxWordLength=N</code> to override the default and expand it to the size you are using for the longest words in your index.</p> <h3 id="resources">Resources and examples</h3> <p> If you want to use your Trix index in a track hub, see the <code>searchTrix</code> setting in the <a href="trackDb/trackDbHub.html#searchTrix">Track Database Definition Document</a>.</p> <p>