Foldseek driver + Structure base Pangenome

.gitignore

@@ -32,4 +32,7 @@ anvio/data/misc/SCG_TAXONOMY/GTDB/SCG_SEARCH_DATABASES/*.dmnd
 anvio/tests/sandbox/update-contigs-and-bams-for-mini-test/output
 anvio/tests/sandbox/test_visualize_split_coverages/TEST_OUTDIR
 anvio/data/misc/KEGG/
+anvio/data/misc/PROSTT5/
 anvio/data/misc/TRNA_TAXONOMY/
+anvio/data/interactive/node_modules
+anvio/data/interactive/package-lock.json

anvio/__init__.py

@@ -3687,7 +3687,46 @@ def TABULATE(table, header, numalign="right", max_width=0):
                      "a comma-separated list. The default stats are 'detection' and "
                      "'mean_coverage_Q2Q3'. To see a list of available stats, use this flag "
                      "and provide an absolutely ridiculous string after it (we suggest 'cattywampus', but you do you)."}
-    )
+    ),
+    'pan-mode': (
+            ['--pan-mode'],
+            {'default': constants.pangenome_mode_default,
+             'choices': constants.pangenome_modes_available,
+             'type': str,
+             'help': f"Computation mode for the pangenome. Default mode is '{constants.pangenome_mode_default}'."}
+    ),
+   'prostt5-data-dir': (
+            ['--prostt5-data-dir'],
+            {'default': None,
+             'type': str,
+             'metavar': 'PATH',
+             'help': "The path for the PROSTT5 Weights to be stored. "
+                     "If you leave it as is without specifying anything, anvi'o will set up everything in "
+                     "a pre-defined default directory. The advantage of using "
+                     "the default directory at the time of set up is that every user of anvi'o on a computer "
+                     "system will be using a single data directory, but then you may need to run the setup "
+                     "program with superuser privileges. If you don't have superuser privileges, then you can "
+                     "use this parameter to tell anvi'o the location you wish to use to setup your weights."}
+        ),
+   'foldseek-search-results': (
+            ['--foldseek-search-results'],
+            {'default': None,
+             'type': str,
+             'metavar': 'FILE PATH',
+             'help': "File path for foldseek search results generated by anvi'o during a previous run. When "
+                     "a file path is provided, anvi'o will try to utilize that output file rather than "
+                     "re-running foldseek. This is mostly for debugging purposes and we strongly advice "
+                     "you to not use it unless you consider yourself to be a hacker-type."}
+        ),
+   'init-pan-mode': (
+            ['--init-pan-mode'],
+            {'default': constants.pangenome_mode_default,
+             'choices': constants.pangenome_modes_available,
+            'type': str,
+            'help': "Specify the structure type to include additional data tables for the Structural Pangenome summary. "
+                     "This option allows you to enhance the analysis by integrating specific structural data, "
+                     "which can provide deeper insights into the pangenomic relationships among the samples."}
+        ),
 }
 
 # two functions that works with the dictionary above.

anvio/bottleroutes.py

@@ -196,6 +196,8 @@ def register_routes(self):
         self.route('/data/get_gene_info/<gene_callers_id>',    callback=self.get_gene_info)
         self.route('/data/get_metabolism',                     callback=self.get_metabolism)
         self.route('/data/get_scale_bar',                      callback=self.get_scale_bar, method='POST')
+        self.route('/data/get_psgc_data/<psgc_name>',          callback=self.get_psgc_data)
+        self.route('/data/get_psgc_type_data/<psgc_name>',    callback=self.get_psgc_type_data)
 
 
     def run_application(self, ip, port):
@@ -1434,6 +1436,55 @@ def get_additional_gc_data(self, gc_id, gc_key):
         except:
             return json.dumps({'status': 1})
 
+
+    def get_psgc_data(self, psgc_name):
+        """Gets PSGC data for the inspection page"""
+
+        try:
+            psgc_data = {
+                psgc_name: []
+            }
+
+            associated_gcs = []
+            for gc_id, psgc_id in self.interactive.gc_psgc_associations.items():
+                if psgc_id == psgc_name:
+                    associated_gcs.append(gc_id)
+
+            for gene_id, gene_entries in self.interactive.gc_tracker.items():
+                for gene_data in gene_entries:
+                    gc_id = gene_data['gene_cluster_id']
+
+                    # Skip if this gene's cluster is not associated with our PSGC
+                    # Its important to do this before we add the gene info to psgc_data
+                    if gc_id not in associated_gcs:
+                        continue
+
+                    gene_info = {
+                        'gene_callers_id': gene_id,
+                        'gene_cluster_id': gc_id,
+                        'genome_name': gene_data['genome_name'],
+                        'alignment_summary': gene_data['alignment_summary']
+                    }
+                    psgc_data[psgc_name].append(gene_info)
+
+            return json.dumps({'status': 0, 'data': psgc_data})
+
+        except Exception as e:
+            return json.dumps({'status': 1, 'message': f"Error getting PSGC data: {str(e)}"})
+
+
+    def get_psgc_type_data(self, psgc_name):
+        """Gets GC type classification data for PSGC (core/singleton/accessory)"""
+        try:
+            gc_types = {}
+            if psgc_name in self.interactive.items_additional_data_dict:
+                gc_types = json.loads(self.interactive.items_additional_data_dict[psgc_name].get('gc_types', '{}'))
+
+            return json.dumps({'status': 0, 'data': {psgc_name: gc_types}})
+        except Exception as e:
+            return json.dumps({'status': 1, 'message': f"Error getting PSGC type data: {str(e)}"})
+
+
     def reroot_tree(self):
         # Get the Newick tree string from the form data
         newick = request.forms.get('newick')

anvio/constants.py

@@ -32,12 +32,23 @@
 # and other purposes
 IS_ESSENTIAL_FIELD = lambda f: (not f.startswith('__')) and (f not in ["contig", "GC_content", "length"])
 
+# MODELLER / PDF file paths
 default_pdb_database_path = os.path.join(os.path.dirname(anvio.__file__), 'data/misc/PDB.db')
 default_modeller_database_dir = os.path.join(os.path.dirname(anvio.__file__), 'data/misc/MODELLER/db')
 default_modeller_scripts_dir = os.path.join(os.path.dirname(anvio.__file__), 'data/misc/MODELLER/scripts')
 
+# Interacdome data path
 default_interacdome_data_path = os.path.join(os.path.dirname(anvio.__file__), 'data/misc/Interacdome')
 
+# ProstT5 data files
+default_prostt5_weight_path = os.path.join(os.path.dirname(anvio.__file__), 'data/misc/PROSTT5/weights')
+
+# Pangenome mode trackers
+PAN_STRUCTURE_MODE = 'structure-informed'
+PAN_SEQUENCE_MODE = 'sequence-based'
+pangenome_modes_available = [PAN_SEQUENCE_MODE, PAN_STRUCTURE_MODE]
+pangenome_mode_default = PAN_SEQUENCE_MODE
+
 clustering_configs_dir = os.path.join(os.path.dirname(anvio.__file__), 'data/clusterconfigs')
 clustering_configs = {}
 

anvio/data/interactive/css/geneclusters.css

@@ -1,4 +1,5 @@
 @import url('normalize.css');
+@import url('variable.css');
 
 #chart-container {
     width: 1200px;
@@ -227,6 +228,48 @@ td {
     padding-bottom: 10px;
 }
 
+.type-core {
+    fill: var(--success-600);
+}
+
+.type-accessory {
+    fill: var(--warning-600);
+}
+
+.type-singleton {
+    fill: var(--info-600);
+}
+
+#display-search-controls {
+    padding: 12px;
+    margin-left: auto;
+    margin-right: auto;
+    background-color: #f5f5f5;
+    border-radius: 9px;
+    margin-bottom: 17px;
+}
+
+.btn-link {
+    background: none;
+    border: none;
+    padding: 0;
+    color: inherit;
+    font-size: 1.5rem;
+    cursor: pointer;
+}
+
+.btn-link:hover {
+    color: var(--warning-600);
+}
+
+.display-text {
+    font-size: 1rem;
+}
+
+#info-icon {
+    font-size: 1.2rem;
+}
+
 @media (min-width: 768px) {
     .gc-container {
         padding: 0 2rem;

anvio/data/interactive/geneclusters.html

@@ -10,6 +10,7 @@
         <meta name="google" value="notranslate">
         <link rel="stylesheet" type="text/css" href="css/loading.css" />
         <link rel="stylesheet" href="./node_modules/bootstrap/dist/css/bootstrap.css" />
+        <link rel="stylesheet" href="./node_modules/bootstrap-icons/font/bootstrap-icons.min.css" />
         <link rel="stylesheet" type="text/css" href="css/geneclusters.css" />
         <link rel="stylesheet" href="./node_modules/toastr/build/toastr.min.css" type="text/css" />
         <link href='https://fonts.googleapis.com/css?family=Lato:300,700' rel='stylesheet' type='text/css' />
@@ -40,19 +41,45 @@ <h1 id="header"></h1>
 
             <section class="main">
                 <div id="display-controls">
-                    Wrap: <input type="text" id="wrap_length" value="1042" size="5" onchange="createDisplay(); "/>
-                          <input type="button" value="-" onclick="increaseVal('wrap_length', -5); createDisplay();" />
-                          <input type="button" value="+" onclick="increaseVal('wrap_length', 5); createDisplay();" />
-                          <span style="display: inline-block; width: 20px;">&nbsp;</span>
-                    Font Size: <input type="text" id="font_size" value="12" size="5" onchange="createDisplay(); "/>
-                          <input type="button" value="-" onclick="increaseVal('font_size', -1); createDisplay();" />
-                          <input type="button" value="+" onclick="increaseVal('font_size', 1); createDisplay();" />
-                          <input type="button" value="Color Settings" onclick="$('#display-conservation-controls').toggle(); " />
+                    <span class="display-text">Wrap:</span>
+                    <input type="text"  class="btn btn-outline-secondary btn-sm" id="wrap_length" value="1042" size="4" onchange="createDisplay(); "/>
+                    <input type="button" class="btn btn-outline-secondary btn-sm" value="-" onclick="increaseVal('wrap_length', -5); createDisplay();" />
+                    <input type="button" class="btn btn-outline-secondary btn-sm" value="+" onclick="increaseVal('wrap_length', 5); createDisplay();" />
+                    <span style="display: inline-block; width: 20px;">&nbsp;</span>
+                    <span class="display-text">Font Size:</span>  
+                    <input type="text" class="btn btn-outline-secondary btn-sm" id="font_size" value="12" size="2" onchange="createDisplay(); "/>
+                    <input type="button" class="btn btn-outline-secondary btn-sm" value="-" onclick="increaseVal('font_size', -1); createDisplay();" />
+                    <input type="button" class="btn btn-outline-secondary btn-sm" value="+" onclick="increaseVal('font_size', 1); createDisplay();" />
+                    <input type="button" class="btn btn-outline-secondary btn-sm" value="Color Settings" onclick="$('#display-conservation-controls').toggle(); " />
+                    <input type="button" class="btn btn-outline-secondary btn-sm" value="Filter" onclick="$('#display-search-controls').toggle();" />
                 </div>
 
                 <div class="d-flex justify-content-center">
                     <div class="col-6" id="display-conservation-controls" style="display: none"></div>
                 </div>
+                <div class="d-flex justify-content-center align-items-center">
+                    <div class="col-4" id="display-search-controls" style="display: none">
+                        <div class="d-flex justify-content-center">
+                            <div class="d-flex">
+                                <div class="d-flex justify-content-center align-items-center mr-3">
+                                    <span type="button" class="btn btn-link" data-toggle="modal" data-target="#infoModal" data-help="info">
+                                        <span id="info-icon" class="bi bi-info-circle" aria-hidden="true"></span>
+                                    </span>
+                                </div>
+                                <div class="d-flex">
+                                    <div class="input-group">
+                                        <span class="input-group-text" style="cursor: pointer;" onclick="filterGenome();">
+                                          <svg xmlns="http://www.w3.org/2000/svg" width="16" height="16" fill="currentColor" class="bi bi-search" viewBox="0 0 16 16">
+                                                <path d="M11.742 10.344a6.5 6.5 0 1 0-1.397 1.398h-.001q.044.06.098.115l3.85 3.85a1 1 0 0 0 1.415-1.414l-3.85-3.85a1 1 0 0 0-.115-.1zM12 6.5a5.5 5.5 0 1 1-11 0 5.5 5.5 0 0 1 11 0"></path>
+                                          </svg>
+                                        </span>
+                                        <input type="text" id="filter-genome" class="form-control d-flex" placeholder="Filter Genomes" onchange="filterGenome(); "/>
+                                    </div>
+                                </div>
+                            </div>
+                        </div>
+                    </div>
+                </div>
                 <div id="display-gcc-acc" class="d-flex justify-content-center">
                     <div id="display-gcc-inner-div">
                         <table id="gc-acc-main" class="table table-responsive col-12">
@@ -80,5 +107,30 @@ <h1 id="header"></h1>
                 $('#' + id).val(Math.max(1,parseInt($('#' + id).val()) + delta));
             }
         </script>
+
+        <!-- Searc GC Modal -->
+        <div class="modal fade" id="infoModal" tabindex="-1" role="dialog" aria-labelledby="infoModalLabel" aria-hidden="true">
+            <div class="modal-dialog" role="document">
+                <div class="modal-content">
+                    <div class="modal-header">
+                        <h4 class="modal-title" id="infoModalLabel">Genome Search</h4>
+                        <button type="button" class="close" data-dismiss="modal" aria-label="Close">
+                            <span aria-hidden="true">&times;</span>
+                        </button>
+                    </div>
+                    <div class="modal-body">
+                        <p>To use the genome search, you can enter one or more genome names separated by commas. For example:</p>
+                        <ul>
+                            <li>Single entry: <code>GENOME_NAME1</code></li>
+                            <li>Multiple entries: <code>GENOME_NAME1, GENOME_NAME2</code></li>
+                        </ul>
+                        <p>This allows you to search for specific genomes in the dataset.</p>
+                    </div>
+                    <div class="modal-footer">
+                        <button type="button" class="btn btn-secondary" data-dismiss="modal">Close</button>
+                    </div>
+                </div>
+            </div>
+        </div>
     </body>
 </html>

anvio/data/interactive/js/bin.js

@@ -76,24 +76,24 @@ Bins.prototype.NewBin = function(id, binState) {
 
     var template = `<tr bin-id="${id}" class="bin-row">
                        <td><input type="radio" name="active_bin" value="${id}"></td>
-                       <td><div id="bin_color_${id}" class="colorpicker" color="${color}" style="background-color: ${color}"></td>
+                       <td><div id="bin_color_${id}" class="colorpicker" color="${color}" style="background-color: ${color}"></div></td>
                        <td data-value="${name}">
                             <input type="text" class="bin-name" oninput="this.value = event.target.value.replaceAll(' ', '_');" onChange="emit('bin-settings-changed'); this.parentNode.setAttribute('data-value', this.value);" size="21" id="bin_name_${id}" value="${name}">
                         </td>
-                       ${mode != 'pan' ? `
+                       ${mode !== 'pan' && mode !== 'structure' ? `
                            <td data-value="${contig_count}" class="num-items"><input type="button" value="${contig_count}" title="Click for contig names" onClick="showContigNames(${id});"></td>
                            <td data-value="${contig_length}" class="length-sum"><span>${contig_length}</span></td>
                        ` : ''}
-                       ${mode == 'pan' ? `
+                       ${mode === 'pan' || mode === 'structure' ? `
                             <td data-value="${num_gene_clusters}" class="num-gene-clusters"><input type="button" value="${num_gene_clusters}" title="Click for quick gene cluster summaries" onClick="showGeneClusterDetails(${id});"></td>
                             <td data-value="${num_gene_calls}" class="num-gene-calls"><input type="button" value="${num_gene_calls}"></td>
                        ` : `
                             <td data-value="${completeness}" class="completeness"><input type="button" value="${completeness}" title="Click for completeness table" onClick="showCompleteness(${id});"></td>
-                            <td data-value="${redundancy}" class="redundancy"><input type="button" value="${redundancy}" title="Click for redundant hits" onClick="showRedundants(${id}); "></td>
+                            <td data-value="${redundancy}" class="redundancy"><input type="button" value="${redundancy}" title="Click for redundant hits" onClick="showRedundants(${id});"></td>
                        `}
                        <td><center><span class="default-bin-icon bi bi-trash-fill fa-lg" aria-hidden="true" alt="Delete this bin" title="Delete this bin" onClick="bins.DeleteBin(${id});"></span></center></td>
                     </tr>
-                    ${ mode === 'full' || mode === 'refine' || mode === 'manual' || mode === 'pan' ? `<tr style="${ $('#estimate_taxonomy').is(':checked') ? `` : `display: none;`}" data-parent="${id}">
+                    ${ mode === 'full' || mode === 'refine' || mode === 'manual' || mode === 'pan' || mode === 'structure' ? `<tr style="${ $('#estimate_taxonomy').is(':checked') ? `` : `display: none;`}" data-parent="${id}">
                     <td style="border-top: 0px;">&nbsp;</td>
                     <td style="border-top: 0px;">&nbsp;</td>
                     <td colspan="6" style="border-top: 0px; padding-top: 0px;">
@@ -470,7 +470,7 @@ Bins.prototype.UpdateBinsWindow = function(bin_list) {
     for (let i = 0; i < bin_list.length; i++) {
         let bin_id = bin_list[i];
 
-        if (mode == 'pan') {
+        if (mode === 'pan' || mode === 'structure') {
             let num_gene_clusters = 0;
             let num_gene_calls = 0;
 

anvio/data/interactive/js/color-coding.js

@@ -329,6 +329,7 @@ function initializeCheckBoxes(){
 
     var all = document.createElement("button");
     all.innerHTML = "Check All";
+    all.className = "btn btn-outline-secondary btn-sm";
     all.style = "margin-right:20px;"
     all.onclick  = (function() {
         for (i = 0; i < labels.length; i++){
@@ -341,6 +342,7 @@ function initializeCheckBoxes(){
 
     var none = document.createElement("button")
     none.innerHTML = "Uncheck All"
+    none.className = "btn btn-outline-secondary btn-sm";
     none.onclick  = (function() {
         for (i = 0; i < labels.length; i++){
             var letter = String(labels[i]);