Add sequence query functionnality

GuillaumeHolley · Nov 21, 2016 · f671eb9 · f671eb9
1 parent f2c1e4a
commit f671eb9
Show file tree

Hide file tree

Showing 217 changed files with 1,353 additions and 1,197 deletions.
diff --git a/.gitignore b/.gitignore
diff --git a/Doxyfile b/Doxyfile
diff --git a/LICENSE b/LICENSE
diff --git a/Makefile.in b/Makefile.in
diff --git a/README.md b/README.md
@@ -110,6 +110,7 @@ bft build k treshold_compression {kmers|kmers_comp} list_genome_files output_fil
 bft load file_bft [-add_genomes {kmers|kmers_comp} list_genome_files output_file] [Options]
 
 Options:
+[-query_sequences threshold list_sequence_files]
 [-query_kmers {kmers|kmers_comp} list_kmer_files]
 [-query_branching {kmers|kmers_comp} list_kmer_files]
 [-extract_kmers {kmers|kmers_comp} kmers_file]
@@ -133,6 +134,7 @@ Command **load** loads a BFT from file *file_bft*.
 ### Options
 
 * **-add_genomes** adds the genomes listed in *list_genome_files* to the BFT stored in *file_bft*, the new BFT is written in *output_file*
+* **-query_sequences** queries the BFT for the sequences written in the files of *list_sequence_files*. For each file of *list_sequence_files* is output a CSV file: columns are the genomes represented in the BFT, rows are the queried sequences, the intersection of a column and a row is a binary value indicating if the sequence represented by the row is present in the genome represented by the column. Threshold is a float (0 < threshold <= 1) indicating the percentage of *k*-mers from each query sequence that must occur in sample *x* to be reported present in sample *x*.
 * **-query_kmers** queries the BFT for *k*-mers written in the files of *list_kmer_files*. For each file of *list_kmer_files* is output a CSV file: columns are the genomes represented in the BFT, rows are the queried *k*-mers, the intersection of a column and a row is a binary value indicating if the *k*-mer represented by the row is present in the genome represented by the column.
 * **-query_branching** queries the BFT for the number of *k*-mers written in the files of *list_kmer_files* that are branching in the colored de-Bruijn graph represented by the BFT.
 * **-extract_kmers** extracts the *k*-mers stored in the BFT and writes them to a *k*-mers file named *kmers_file* (see below for input file types).

diff --git a/configure b/configure
@@ -650,6 +650,7 @@ infodir
 docdir
 oldincludedir
 includedir
+runstatedir
 localstatedir
 sharedstatedir
 sysconfdir
@@ -721,6 +722,7 @@ datadir='${datarootdir}'
 sysconfdir='${prefix}/etc'
 sharedstatedir='${prefix}/com'
 localstatedir='${prefix}/var'
+runstatedir='${localstatedir}/run'
 includedir='${prefix}/include'
 oldincludedir='/usr/include'
 docdir='${datarootdir}/doc/${PACKAGE_TARNAME}'
@@ -973,6 +975,15 @@ do
   | -silent | --silent | --silen | --sile | --sil)
     silent=yes ;;
 
+  -runstatedir | --runstatedir | --runstatedi | --runstated \
+  | --runstate | --runstat | --runsta | --runst | --runs \
+  | --run | --ru | --r)
+    ac_prev=runstatedir ;;
+  -runstatedir=* | --runstatedir=* | --runstatedi=* | --runstated=* \
+  | --runstate=* | --runstat=* | --runsta=* | --runst=* | --runs=* \
+  | --run=* | --ru=* | --r=*)
+    runstatedir=$ac_optarg ;;
+
   -sbindir | --sbindir | --sbindi | --sbind | --sbin | --sbi | --sb)
     ac_prev=sbindir ;;
   -sbindir=* | --sbindir=* | --sbindi=* | --sbind=* | --sbin=* \
@@ -1110,7 +1121,7 @@ fi
 for ac_var in	exec_prefix prefix bindir sbindir libexecdir datarootdir \
 		datadir sysconfdir sharedstatedir localstatedir includedir \
 		oldincludedir docdir infodir htmldir dvidir pdfdir psdir \
-		libdir localedir mandir
+		libdir localedir mandir runstatedir
 do
   eval ac_val=\$$ac_var
   # Remove trailing slashes.
@@ -1263,6 +1274,7 @@ Fine tuning of the installation directories:
   --sysconfdir=DIR        read-only single-machine data [PREFIX/etc]
   --sharedstatedir=DIR    modifiable architecture-independent data [PREFIX/com]
   --localstatedir=DIR     modifiable single-machine data [PREFIX/var]
+  --runstatedir=DIR       modifiable per-process data [LOCALSTATEDIR/run]
   --libdir=DIR            object code libraries [EPREFIX/lib]
   --includedir=DIR        C header files [PREFIX/include]
   --oldincludedir=DIR     C header files for non-gcc [/usr/include]

diff --git a/doc/doxygen/doxygen_sqlite3.db b/doc/doxygen/doxygen_sqlite3.db
diff --git a/doc/doxygen/html/Node_8h_source.html b/doc/doxygen/html/Node_8h_source.html
diff --git a/doc/doxygen/html/annotated.html b/doc/doxygen/html/annotated.html
@@ -3,7 +3,7 @@
 <head>
 <meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
 <meta http-equiv="X-UA-Compatible" content="IE=9"/>
-<meta name="generator" content="Doxygen 1.8.9.1"/>
+<meta name="generator" content="Doxygen 1.8.11"/>
 <title>Bloom Filter Trie: Data Structures</title>
 <link href="tabs.css" rel="stylesheet" type="text/css"/>
 <script type="text/javascript" src="jquery.js"></script>
@@ -22,7 +22,7 @@
 <table cellspacing="0" cellpadding="0">
  <tbody>
  <tr style="height: 56px;">
-  <td style="padding-left: 0.5em;">
+  <td id="projectalign" style="padding-left: 0.5em;">
    <div id="projectname">Bloom Filter Trie
    </div>
   </td>
@@ -31,7 +31,7 @@
 </table>
 </div>
 <!-- end header part -->
-<!-- Generated by Doxygen 1.8.9.1 -->
+<!-- Generated by Doxygen 1.8.11 -->
 <script type="text/javascript">
 var searchBox = new SearchBox("searchBox", "search",false,'Search');
 </script>
@@ -95,9 +95,9 @@
 </div><!-- contents -->
 <!-- start footer part -->
 <hr class="footer"/><address class="footer"><small>
-Generated on Wed Nov 9 2016 17:22:01 for Bloom Filter Trie by &#160;<a href="http://www.doxygen.org/index.html">
+Generated on Mon Nov 21 2016 17:45:16 for Bloom Filter Trie by &#160;<a href="http://www.doxygen.org/index.html">
 <img class="footer" src="doxygen.png" alt="doxygen"/>
-</a> 1.8.9.1
+</a> 1.8.11
 </small></address>
 </body>
 </html>
diff --git a/doc/doxygen/html/arrowdown.png b/doc/doxygen/html/arrowdown.png
diff --git a/doc/doxygen/html/arrowright.png b/doc/doxygen/html/arrowright.png
diff --git a/doc/doxygen/html/bc_s.png b/doc/doxygen/html/bc_s.png
diff --git a/doc/doxygen/html/bdwn.png b/doc/doxygen/html/bdwn.png
diff --git a/doc/doxygen/html/bft_8h.html b/doc/doxygen/html/bft_8h.html
@@ -3,7 +3,7 @@
 <head>
 <meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
 <meta http-equiv="X-UA-Compatible" content="IE=9"/>
-<meta name="generator" content="Doxygen 1.8.9.1"/>
+<meta name="generator" content="Doxygen 1.8.11"/>
 <title>Bloom Filter Trie: include/bft.h File Reference</title>
 <link href="tabs.css" rel="stylesheet" type="text/css"/>
 <script type="text/javascript" src="jquery.js"></script>
@@ -22,7 +22,7 @@
 <table cellspacing="0" cellpadding="0">
  <tbody>
  <tr style="height: 56px;">
-  <td style="padding-left: 0.5em;">
+  <td id="projectalign" style="padding-left: 0.5em;">
    <div id="projectname">Bloom Filter Trie
    </div>
   </td>
@@ -31,7 +31,7 @@
 </table>
 </div>
 <!-- end header part -->
-<!-- Generated by Doxygen 1.8.9.1 -->
+<!-- Generated by Doxygen 1.8.11 -->
 <script type="text/javascript">
 var searchBox = new SearchBox("searchBox", "search",false,'Search');
 </script>
@@ -192,12 +192,6 @@
 <tr class="memitem:ac5c1b6b0b04abfdf205f966299f042fd"><td class="memItemLeft" align="right" valign="top">void&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#ac5c1b6b0b04abfdf205f966299f042fd">free_BFT_kmer_content</a> (<a class="el" href="structBFT__kmer.html">BFT_kmer</a> *bft_kmer, int nb_bft_kmer)</td></tr>
 <tr class="memdesc:ac5c1b6b0b04abfdf205f966299f042fd"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function freeing the content of allocated BFT_kmers.  <a href="#ac5c1b6b0b04abfdf205f966299f042fd">More...</a><br /></td></tr>
 <tr class="separator:ac5c1b6b0b04abfdf205f966299f042fd"><td class="memSeparator" colspan="2">&#160;</td></tr>
-<tr class="memitem:a67f15f4eeebe6fe620c467f8180e1f31"><td class="memItemLeft" align="right" valign="top"><a class="el" href="structBFT__kmer.html">BFT_kmer</a> *&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#a67f15f4eeebe6fe620c467f8180e1f31">get_kmer</a> (const char *kmer, <a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *bft)</td></tr>
-<tr class="memdesc:a67f15f4eeebe6fe620c467f8180e1f31"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function searching for a k-mer in a BFT.  <a href="#a67f15f4eeebe6fe620c467f8180e1f31">More...</a><br /></td></tr>
-<tr class="separator:a67f15f4eeebe6fe620c467f8180e1f31"><td class="memSeparator" colspan="2">&#160;</td></tr>
-<tr class="memitem:a170137e867f0ba24423badc9e1db872b"><td class="memItemLeft" align="right" valign="top">bool&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#a170137e867f0ba24423badc9e1db872b">is_kmer_in_cdbg</a> (<a class="el" href="structBFT__kmer.html">BFT_kmer</a> *bft_kmer)</td></tr>
-<tr class="memdesc:a170137e867f0ba24423badc9e1db872b"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function testing if a k-mer is in a BFT.  <a href="#a170137e867f0ba24423badc9e1db872b">More...</a><br /></td></tr>
-<tr class="separator:a170137e867f0ba24423badc9e1db872b"><td class="memSeparator" colspan="2">&#160;</td></tr>
 <tr class="memitem:ab9b8daf6a5685786083c258526ef87cb"><td class="memItemLeft" align="right" valign="top">void&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#ab9b8daf6a5685786083c258526ef87cb">extract_kmers_to_disk</a> (<a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *bft, char *filename_output, bool compressed_output)</td></tr>
 <tr class="memdesc:ab9b8daf6a5685786083c258526ef87cb"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function extracting the k-mers of a BFT in a file.  <a href="#ab9b8daf6a5685786083c258526ef87cb">More...</a><br /></td></tr>
 <tr class="separator:ab9b8daf6a5685786083c258526ef87cb"><td class="memSeparator" colspan="2">&#160;</td></tr>
@@ -207,6 +201,24 @@
 <tr class="memitem:a77a802e5edc240a7a6a705ef6593799d"><td class="memItemLeft" align="right" valign="top">size_t&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#a77a802e5edc240a7a6a705ef6593799d">write_kmer_comp_to_disk</a> (<a class="el" href="structBFT__kmer.html">BFT_kmer</a> *bft_kmer, <a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *bft, va_list args)</td></tr>
 <tr class="memdesc:a77a802e5edc240a7a6a705ef6593799d"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function writing an 2 bits encoded k-mer in a file.  <a href="#a77a802e5edc240a7a6a705ef6593799d">More...</a><br /></td></tr>
 <tr class="separator:a77a802e5edc240a7a6a705ef6593799d"><td class="memSeparator" colspan="2">&#160;</td></tr>
+<tr><td colspan="2"><div class="groupHeader">Query functions</div></td></tr>
+<tr><td colspan="2"><div class="groupText"><p>These functions query for k-mers or sequences. </p>
+</div></td></tr>
+<tr class="memitem:a67f15f4eeebe6fe620c467f8180e1f31"><td class="memItemLeft" align="right" valign="top"><a class="el" href="structBFT__kmer.html">BFT_kmer</a> *&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#a67f15f4eeebe6fe620c467f8180e1f31">get_kmer</a> (const char *kmer, <a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *bft)</td></tr>
+<tr class="memdesc:a67f15f4eeebe6fe620c467f8180e1f31"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function searching for a k-mer in a BFT.  <a href="#a67f15f4eeebe6fe620c467f8180e1f31">More...</a><br /></td></tr>
+<tr class="separator:a67f15f4eeebe6fe620c467f8180e1f31"><td class="memSeparator" colspan="2">&#160;</td></tr>
+<tr class="memitem:a170137e867f0ba24423badc9e1db872b"><td class="memItemLeft" align="right" valign="top">bool&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#a170137e867f0ba24423badc9e1db872b">is_kmer_in_cdbg</a> (<a class="el" href="structBFT__kmer.html">BFT_kmer</a> *bft_kmer)</td></tr>
+<tr class="memdesc:a170137e867f0ba24423badc9e1db872b"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function testing if a k-mer is in a BFT.  <a href="#a170137e867f0ba24423badc9e1db872b">More...</a><br /></td></tr>
+<tr class="separator:a170137e867f0ba24423badc9e1db872b"><td class="memSeparator" colspan="2">&#160;</td></tr>
+<tr class="memitem:aa6c30881d62f797a766d5ad9e1548b21"><td class="memItemLeft" align="right" valign="top">uint32_t *&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#aa6c30881d62f797a766d5ad9e1548b21">query_sequence</a> (<a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *bft, char *sequence, double threshold)</td></tr>
+<tr class="memdesc:aa6c30881d62f797a766d5ad9e1548b21"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function querying a BFT for a sequence.  <a href="#aa6c30881d62f797a766d5ad9e1548b21">More...</a><br /></td></tr>
+<tr class="separator:aa6c30881d62f797a766d5ad9e1548b21"><td class="memSeparator" colspan="2">&#160;</td></tr>
+<tr><td colspan="2"><div class="groupHeader">Pattern matching functions</div></td></tr>
+<tr><td colspan="2"><div class="groupText"><p>These functions provide pattern matching functionalities over the k-mers or paths of a colored de Bruijn graph stored as a BFT. </p>
+</div></td></tr>
+<tr class="memitem:a6f93be4a2e37c7355c79079e8c38530a"><td class="memItemLeft" align="right" valign="top">bool&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#a6f93be4a2e37c7355c79079e8c38530a">prefix_matching</a> (<a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *bft, char *prefix, <a class="el" href="bft_8h.html#aac9ee38523fda7ab4f8eb1238e98d315">BFT_func_ptr</a> f,...)</td></tr>
+<tr class="memdesc:a6f93be4a2e37c7355c79079e8c38530a"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function for prefix matching over the k-mers of a BFT.  <a href="#a6f93be4a2e37c7355c79079e8c38530a">More...</a><br /></td></tr>
+<tr class="separator:a6f93be4a2e37c7355c79079e8c38530a"><td class="memSeparator" colspan="2">&#160;</td></tr>
 <tr><td colspan="2"><div class="groupHeader">Marking functions</div></td></tr>
 <tr><td colspan="2"><div class="groupText"><p>These functions allow to mark k-mers of a colored de Bruijn graph with flags. </p>
 </div></td></tr>
@@ -249,12 +261,6 @@
 <tr class="memitem:aa616e55874abfd5811c8854586718c35"><td class="memItemLeft" align="right" valign="top">void&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#aa616e55874abfd5811c8854586718c35">v_iterate_over_kmers</a> (<a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *bft, <a class="el" href="bft_8h.html#aac9ee38523fda7ab4f8eb1238e98d315">BFT_func_ptr</a> f, va_list args)</td></tr>
 <tr class="memdesc:aa616e55874abfd5811c8854586718c35"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function iterating over the k-mers of a BFT.  <a href="#aa616e55874abfd5811c8854586718c35">More...</a><br /></td></tr>
 <tr class="separator:aa616e55874abfd5811c8854586718c35"><td class="memSeparator" colspan="2">&#160;</td></tr>
-<tr><td colspan="2"><div class="groupHeader">Pattern matching functions</div></td></tr>
-<tr><td colspan="2"><div class="groupText"><p>These functions provide pattern matching functionalities over the k-mers or paths of a colored de Bruijn graph stored as a BFT. </p>
-</div></td></tr>
-<tr class="memitem:a6f93be4a2e37c7355c79079e8c38530a"><td class="memItemLeft" align="right" valign="top">bool&#160;</td><td class="memItemRight" valign="bottom"><a class="el" href="bft_8h.html#a6f93be4a2e37c7355c79079e8c38530a">prefix_matching</a> (<a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *bft, char *prefix, <a class="el" href="bft_8h.html#aac9ee38523fda7ab4f8eb1238e98d315">BFT_func_ptr</a> f,...)</td></tr>
-<tr class="memdesc:a6f93be4a2e37c7355c79079e8c38530a"><td class="mdescLeft">&#160;</td><td class="mdescRight">Function for prefix matching over the k-mers of a BFT.  <a href="#a6f93be4a2e37c7355c79079e8c38530a">More...</a><br /></td></tr>
-<tr class="separator:a6f93be4a2e37c7355c79079e8c38530a"><td class="memSeparator" colspan="2">&#160;</td></tr>
 <tr><td colspan="2"><div class="groupHeader">Disk I/O functions</div></td></tr>
 <tr><td colspan="2"><div class="groupText"><p>These functions write and load a BFT from disk. </p>
 </div></td></tr>
@@ -1224,6 +1230,49 @@ <h2 class="groupheader">Function Documentation</h2>
 </dl>
 <dl class="section return"><dt>Returns</dt><dd>a boolean indicating the presence (true) or absence (false) of the k-mer in a the genome. </dd></dl>
 
+</div>
+</div>
+<a class="anchor" id="aa6c30881d62f797a766d5ad9e1548b21"></a>
+<div class="memitem">
+<div class="memproto">
+      <table class="memname">
+        <tr>
+          <td class="memname">uint32_t* query_sequence </td>
+          <td>(</td>
+          <td class="paramtype"><a class="el" href="bft_8h.html#a2db0a80fe662d044625641531e5ce292">BFT</a> *&#160;</td>
+          <td class="paramname"><em>bft</em>, </td>
+        </tr>
+        <tr>
+          <td class="paramkey"></td>
+          <td></td>
+          <td class="paramtype">char *&#160;</td>
+          <td class="paramname"><em>sequence</em>, </td>
+        </tr>
+        <tr>
+          <td class="paramkey"></td>
+          <td></td>
+          <td class="paramtype">double&#160;</td>
+          <td class="paramname"><em>threshold</em>&#160;</td>
+        </tr>
+        <tr>
+          <td></td>
+          <td>)</td>
+          <td></td><td></td>
+        </tr>
+      </table>
+</div><div class="memdoc">
+
+<p>Function querying a BFT for a sequence. </p>
+<dl class="params"><dt>Parameters</dt><dd>
+  <table class="params">
+    <tr><td class="paramname">bft</td><td>is a BFT to be queried. </td></tr>
+    <tr><td class="paramname">sequence</td><td>is a string to query. </td></tr>
+    <tr><td class="paramname">threshold</td><td>is a float (0 &lt; threshold &lt;= 1) indicating the minimum percentage of k-mers from the queried sequence that must be present in a genome to have the queried sequence reported present in this genome. </td></tr>
+  </table>
+  </dd>
+</dl>
+<dl class="section return"><dt>Returns</dt><dd>a pointer to a sorted array of genome identifiers in which the queried sequence occurs (according to parameter threshold) or NULL if the queried sequence is not present in at least one genome (according to parameter threshold). The first element of the array (position 0) indicates how many ids are in this array. </dd></dl>
+
 </div>
 </div>
 <a class="anchor" id="aec4a612e2beec69b717be9e972fa16cb"></a>
@@ -1621,9 +1670,9 @@ <h2 class="groupheader">Function Documentation</h2>
 </div><!-- contents -->
 <!-- start footer part -->
 <hr class="footer"/><address class="footer"><small>
-Generated on Wed Nov 9 2016 17:22:01 for Bloom Filter Trie by &#160;<a href="http://www.doxygen.org/index.html">
+Generated on Mon Nov 21 2016 17:45:16 for Bloom Filter Trie by &#160;<a href="http://www.doxygen.org/index.html">
 <img class="footer" src="doxygen.png" alt="doxygen"/>
-</a> 1.8.9.1
+</a> 1.8.11
 </small></address>
 </body>
 </html>