--- a/filter.xml	Wed Sep 22 21:04:52 2021 +0000
+++ b/filter.xml	Wed Jul 31 18:06:35 2024 +0000
@@ -1,9 +1,9 @@
-<tool id="scanpy_filter" name="Filter" version="@galaxy_version@" profile="@profile@">
+<tool id="scanpy_filter" name="Filter" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="@profile@">
     <description>with scanpy</description>
-    <expand macro="bio_tools"/>
     <macros>
         <import>macros.xml</import>
     </macros>
+    <expand macro="bio_tools"/>
     <expand macro="requirements"/>
     <expand macro="version_command"/>
     <command detect_errors="exit_code"><![CDATA[
@@ -45,14 +45,13 @@
 #else if $method.method == 'tl.filter_rank_genes_groups'
 sc.tl.filter_rank_genes_groups(
     adata,
-    #if str($method.key) != ''
+    #if $method.key
     key='$method.key',
     #end if
-    #if str($method.groupby) != ''
+    #if $method.groupby
     groupby='$method.groupby',
     #end if
     use_raw=$method.use_raw,
-    log=$method.log,
     key_added='$method.key_added',
     min_in_group_fraction=$method.min_in_group_fraction,
     max_out_group_fraction=$method.max_out_group_fraction,
@@ -63,17 +62,11 @@
     adata=adata,
     flavor='$method.flavor.flavor',
     #if $method.flavor.flavor == 'seurat'
-        #if str($method.flavor.min_mean) != ''
-    min_mean=$method.flavor.min_mean,
-        #end if
-        #if str($method.flavor.max_mean) != ''
-    max_mean=$method.flavor.max_mean,
-        #end if
-        #if str($method.flavor.min_disp) != ''
-    min_disp=$method.flavor.min_disp,
-        #end if
+        min_mean=$method.flavor.min_mean,
+        max_mean=$method.flavor.max_mean,
+        min_disp=$method.flavor.min_disp,
         #if str($method.flavor.max_disp) != ''
-    max_disp=$method.flavor.max_disp,
+        max_disp=$method.flavor.max_disp,
         #end if
     #else if $method.flavor.flavor == 'cell_ranger'
     n_top_genes=$method.flavor.n_top_genes,
@@ -105,6 +98,41 @@
     random_state=$method.random_state,
     replace=$method.replace,
     copy=False)
+
+#else if $method.method == "filter_marker"
+
+#if $method.layer_selection.use_raw == 'False':
+    adata.X = adata.layers['$method.layer_selection.layer']
+#end if
+
+def check_marker(adata, group, gene, thresh_mean, thresh_frac, groupby):
+    filtered_data = adata[adata.obs[groupby] == group, adata.var_names == gene]
+    mean_expression = np.mean(filtered_data.X)
+    frac_cell_mean_expression = len(filtered_data.X[filtered_data.X > mean_expression]) / filtered_data.n_obs
+    if ( mean_expression > thresh_mean and frac_cell_mean_expression >= thresh_frac ):
+        return(True)
+    return(False)
+
+header='infer'
+
+#if $method.header == 'not_included':
+    header=None
+#end if
+
+marker_list={key: list(value.values()) for key, value in pd.read_csv('$method.markerfile', sep='\t', index_col=0, header=header).to_dict(orient='index').items()}
+
+for key, value in marker_list.items():
+    marker_list[key] = [x for x in value if check_marker(adata, key, x, $method.thresh_mean, $method.thresh_frac, '$method.groupby')]
+
+# Find the maximum length of lists
+max_len = max(len(lst) for lst in marker_list.values())
+
+# Fill smaller lists with empty values
+for key, value in marker_list.items():
+    marker_list[key] = value + [''] * (max_len - len(value))
+
+df = pd.DataFrame(marker_list).T
+df.to_csv('marker.tsv', sep='\t', index=True)
 #end if
 
 @CMD_anndata_write_outputs@
@@ -120,6 +148,7 @@
                 <option value="pp.highly_variable_genes">Annotate (and filter) highly variable genes, using 'pp.highly_variable_genes'</option>
                 <option value="pp.subsample">Subsample to a fraction of the number of observations, using 'pp.subsample'</option>
                 <option value="pp.downsample_counts">Downsample counts from count matrix, using 'pp.downsample_counts'</option>
+                <option value="filter_marker">Filter markers from count matrix and marker list</option>
             </param>
             <when value="pp.filter_cells">
                 <conditional name="filter">
@@ -173,7 +202,6 @@
                     <expand macro="sanitize_query" />
                 </param>
                 <expand macro="param_use_raw"/>
-                <expand macro="param_log"/>
                 <param argument="key_added" type="text" value="rank_genes_groups_filtered" label="Key that will contain new values">
                     <expand macro="sanitize_query" />
                 </param>
@@ -221,15 +249,40 @@
                 <param argument="random_state" type="integer" value="0" label="Random seed to change subsampling"/>
                 <param argument="replace" type="boolean" truevalue="True" falsevalue="False" checked="false" label="Sample the counts with replacement?"/>
             </when>
+            <when value="filter_marker">
+                <param argument="markerfile" type="data" format="tabular" label="List of markers" help="This should be a tsv where row = group (e.g. celltypes) and columns = markers."></param>
+                <param name="header" type="select" label="Header in the list of markers?">
+                    <option value="included">Header incldued</option>
+                    <option value="not_included">Header not included</option>
+                </param>
+                <param argument="thresh_mean" type="float" min="0.0" value="1.0" label="Minimal average count of all cells of a group (e.g., celltype) for a particular marker" help="Increasing the threshold will result in a smaller marker set."/>
+                <param argument="thresh_frac" type="float" min="0.0" max="1.0" value="0.1" label="Minimal fractions of cells that has a higher count than the average count of all cells of the group for the marker" help="Increasing this threshold might remove marker outliers."/>
+                <conditional name="layer_selection">
+                    <param name="use_raw" type="select" label="Use .X of adata to perform the filtering" help="">
+                        <option value="True">Yes</option>
+                        <option value="False">No</option>
+                    </param>
+                    <when value="False">
+                        <param argument="layer" type="text" value="" label="Key from adata.layers whose value will be used to filter" help="If layers specified then use adata.layers[layer]."/>
+                    </when>
+                    <when value="True"/>
+                </conditional>
+                <param argument="groupby" type="text" value="" label="The key of the observation grouping to consider (e.g., celltype)" help="">
+                    <expand macro="sanitize_query" />
+                </param>
+            </when>
         </conditional>
         <expand macro="inputs_common_advanced"/>
     </inputs>
     <outputs>
         <expand macro="anndata_outputs"/>
+        <data name="marker_out" format="tabular" from_work_dir="marker.tsv" label="${tool.name} on ${on_string}: Markers">
+            <filter>method['method'] == 'filter_marker'</filter>
+        </data>
     </outputs>
     <tests>
-        <test>
-            <!-- test 0 -->
+        <test expect_num_outputs="2">
+            <!-- test 1 -->
             <param name="adata" value="krumsiek11.h5ad" />
             <conditional name="method">
                 <param name="method" value="pp.filter_cells"/>
@@ -252,8 +305,8 @@
             </output>
             <output name="anndata_out" file="pp.filter_cells.krumsiek11-min_counts.h5ad" ftype="h5ad" compare="sim_size"/>
         </test>
-        <test>
-            <!-- test 1 -->
+        <test expect_num_outputs="2">
+            <!-- test 2 -->
             <param name="adata" value="krumsiek11.h5ad" />
             <conditional name="method">
                 <param name="method" value="pp.filter_cells"/>
@@ -274,8 +327,8 @@
             </output>
             <output name="anndata_out" file="pp.filter_cells.krumsiek11-max_genes.h5ad" ftype="h5ad" compare="sim_size"/>
         </test>
-        <test>
-            <!-- test 2 -->
+        <test expect_num_outputs="2">
+            <!-- test 3 -->
             <param name="adata" value="krumsiek11.h5ad" />
             <conditional name="method">
                 <param name="method" value="pp.filter_genes"/>
@@ -295,36 +348,41 @@
             </output>
             <output name="anndata_out" file="pp.filter_genes.krumsiek11-min_counts.h5ad" ftype="h5ad" compare="sim_size"/>
         </test>
-        <!-- <test> -->
-        <!--     <!-\- test 3 -\-> -->
-        <!--     <!-\- Input dataset appears to be missing rank_genes_groups key... -\-> -->
-        <!--     <param name="adata" value="tl.rank_genes_groups.krumsiek11.h5ad" /> -->
-        <!--     <conditional name="method"> -->
-        <!--         <param name="method" value="tl.filter_rank_genes_groups"/> -->
-        <!--         <param name="key" value="rank_genes_groups"/> -->
-        <!--         <param name="use_raw" value="False"/> -->
-        <!--         <param name="log" value="False"/> -->
-        <!--         <param name="key_added" value="rank_genes_groups_filtered"/> -->
-        <!--         <param name="min_in_group_fraction" value="0.25"/> -->
-        <!--         <param name="max_out_group_fraction" value="0.5"/> -->
-        <!--         <param name="min_fold_change" value="3"/> -->
-        <!--     </conditional> -->
-        <!--     <output name="hidden_output"> -->
-        <!--         <assert_contents> -->
-        <!--             <has_text_matching expression="tl.filter_rank_genes_groups"/> -->
-        <!--             <has_text_matching expression="key='rank_genes_groups'"/> -->
-        <!--             <has_text_matching expression="use_raw=False"/> -->
-        <!--             <has_text_matching expression="log=False"/> -->
-        <!--             <has_text_matching expression="key_added='rank_genes_groups_filtered'"/> -->
-        <!--             <has_text_matching expression="min_in_group_fraction=0.25"/> -->
-        <!--             <has_text_matching expression="max_out_group_fraction=0.5"/> -->
-        <!--             <has_text_matching expression="min_fold_change=3"/> -->
-        <!--         </assert_contents> -->
-        <!--     </output> -->
-        <!--     <output name="anndata_out" file="pp.filter_rank_genes_groups.h5ad" ftype="h5ad" compare="sim_size"/> -->
-        <!-- </test> -->
-        <test>
-            <!-- test 4 -->
+
+        <!--  test 4 -->
+        <!-- Fails to write to anndata after tl.filter_rank_genes_groups
+             Issue has been reported here: https://github.com/scverse/anndata/issues/726
+             The current fix is: del adata.uns['rank_genes_groups_filtered']  -->
+        <!--<test expect_num_outputs="2">
+            <param name="adata" value="tl.rank_genes_groups.krumsiek11.h5ad" />
+            <conditional name="method">
+                <param name="method" value="tl.filter_rank_genes_groups"/>
+                <param name="key" value="rank_genes_groups"/>
+                <param name="use_raw" value="False"/>
+                <param name="key_added" value="rank_genes_groups_filtered"/>
+                <param name="min_in_group_fraction" value="0.25"/>
+                <param name="max_out_group_fraction" value="0.5"/>
+                <param name="min_fold_change" value="3"/>
+            </conditional>
+            <section name="advanced_common">
+                <param name="show_log" value="true" />
+            </section>
+            <output name="hidden_output">
+                <assert_contents>
+                    <has_text_matching expression="tl.filter_rank_genes_groups"/>
+                    <has_text_matching expression="key='rank_genes_groups'"/>
+                    <has_text_matching expression="use_raw=False"/>
+                    <has_text_matching expression="log=False"/>
+                    <has_text_matching expression="key_added='rank_genes_groups_filtered'"/>
+                    <has_text_matching expression="min_in_group_fraction=0.25"/>
+                    <has_text_matching expression="max_out_group_fraction=0.5"/>
+                    <has_text_matching expression="min_fold_change=3"/>
+                </assert_contents>
+            </output>
+            <output name="anndata_out" file="pp.filter_rank_genes_groups.h5ad" ftype="h5ad" compare="sim_size"/>
+        </test>-->
+        <test expect_num_outputs="2">
+            <!-- test 5 -->
             <param name="adata" value="blobs.h5ad"/>
             <conditional name="method">
                 <param name="method" value="pp.highly_variable_genes"/>
@@ -353,8 +411,8 @@
             </output>
             <output name="anndata_out" file="pp.highly_variable_genes.seurat.blobs.h5ad" ftype="h5ad" compare="sim_size" delta="100000" delta_frac="0.2"/>
         </test>
-        <test>
-            <!-- test 5 -->
+        <test expect_num_outputs="2">
+            <!-- test 6 -->
             <param name="adata" value="krumsiek11.h5ad" />
             <conditional name="method">
                 <param name="method" value="pp.highly_variable_genes"/>
@@ -379,8 +437,8 @@
             </output>
             <output name="anndata_out" file="pp.highly_variable_genes.krumsiek11-cell_ranger.h5ad" ftype="h5ad" compare="sim_size" delta="100000" delta_frac="0.9"/>
         </test>
-        <test>
-            <!-- test 6 -->
+        <test expect_num_outputs="2">
+            <!-- test 7 -->
             <param name="adata" value="krumsiek11.h5ad" />
             <conditional name="method">
                 <param name="method" value="pp.subsample"/>
@@ -402,8 +460,8 @@
             </output>
             <output name="anndata_out" file="pp.subsample.krumsiek11_fraction.h5ad" ftype="h5ad" compare="sim_size"/>
         </test>
-        <test>
-            <!-- test 7 -->
+        <test expect_num_outputs="2">
+            <!-- test 8 -->
             <param name="adata" value="krumsiek11.h5ad" />
             <conditional name="method">
                 <param name="method" value="pp.subsample"/>
@@ -425,8 +483,8 @@
             </output>
             <output name="anndata_out" file="pp.subsample.krumsiek11_n_obs.h5ad" ftype="h5ad" compare="sim_size"/>
         </test>
-        <test>
-            <!-- test 8 -->
+        <test expect_num_outputs="2">
+            <!-- test 9 -->
             <param name="adata" value="random-randint.h5ad" />
             <conditional name="method">
                 <param name="method" value="pp.downsample_counts"/>
@@ -447,6 +505,32 @@
             </output>
             <output name="anndata_out" file="pp.downsample_counts.random-randint.h5ad" ftype="h5ad" compare="sim_size" delta="10000000" delta_frac="0.5"/>
         </test>
+        <test expect_num_outputs="3">
+            <!-- test 10 -->
+            <param name="adata" value="cosg.rank_genes_groups.newton-cg.pbmc68k_highly_reduced_1.h5ad" />
+            <conditional name="method">
+                <param name="method" value="filter_marker"/>
+                <param name="markerfile" value="tl.rank_genes_groups.newton-cg.pbmc68k_highly_reduced_marker_1.tsv"/>
+                <param name="thresh_mean" value="1.0"/>
+                <param name="thresh_frac" value="0.2"/>
+                <param name="layer_selection" value="True"/>
+                <param name="groupby" value="bulk_labels"/>
+            </conditional>
+            <section name="advanced_common">
+                <param name="show_log" value="true" />
+            </section>
+            <output name="hidden_output">
+                <assert_contents>
+                    <has_text_matching expression="adata, key, x, 1.0, 0.2, 'bulk_labels'"/>
+                </assert_contents>
+            </output>
+            <output name="anndata_out" file="cosg.rank_genes_groups.newton-cg.pbmc68k_highly_reduced_1.h5ad" ftype="h5ad">
+                <assert_contents>
+                    <has_h5_keys keys="obs, var, uns" />
+                </assert_contents>
+            </output>
+            <output name="marker_out" file="tl.rank_genes_groups.newton-cg.pbmc68k_highly_reduced_marker_filtered_1.tsv" ftype="tabular" compare="sim_size"/>
+        </test>
     </tests>
     <help><![CDATA[
 
@@ -461,7 +545,7 @@
 `max_counts`, `max_genes` per call.
 
 More details on the `scanpy documentation
-<https://icb-scanpy.readthedocs-hosted.com/en/@version@/api/scanpy.pp.filter_cells.html>`__
+<https://scanpy.readthedocs.io/en/stable/api/scanpy.pp.filter_cells.html>`__
 
 
 Filter genes based on number of cells or counts (`pp.filter_genes`)
@@ -475,14 +559,14 @@
 `max_counts`, `max_cells` per call.
 
 More details on the `scanpy documentation
-<https://icb-scanpy.readthedocs-hosted.com/en/@version@/api/scanpy.pp.filter_genes.html>`__
+<https://scanpy.readthedocs.io/en/stable/api/scanpy.pp.filter_genes.html>`__
 
 
 Filters out genes based on fold change and fraction of genes expressing the gene within and outside the groupby categories (`tl.filter_rank_genes_groups`)
 ==========================================================================================================================================================
 
 More details on the `scanpy documentation
-<https://icb-scanpy.readthedocs-hosted.com/en/@version@/api/scanpy.tl.filter_rank_genes_groups.html>`__
+<https://scanpy.readthedocs.io/en/stable/api/scanpy.tl.filter_rank_genes_groups.html>`__
 
 
 Annotate highly variable genes (`pp.highly_variable_genes`)
@@ -497,7 +581,7 @@
 ======================================================================
 
 More details on the `scanpy documentation
-<https://icb-scanpy.readthedocs-hosted.com/en/@version@/api/scanpy.pp.subsample.html>`__
+<https://scanpy.readthedocs.io/en/stable/api/scanpy.pp.subsample.html>`__
 
 Downsample counts (`pp.downsample_counts`)
 ==========================================
@@ -505,8 +589,18 @@
 Downsample counts so that each cell has no more than `target_counts`. Cells with fewer counts than `target_counts` are unaffected by this. This
 has been implemented by M. D. Luecken.
 
+
+Filter marker genes (`filter_marker`)
+======================================================================
+
+This option is specific for celltype marker gene detection. You can generate a celltype marker gene file (tsv) with **COSG** provided at Galaxy.
+
+The marker gene file should have as rows celltypes and columns as marker genes. Each celltype can have varying number of marker genes.
+
+A marker gene is returned (retained in the list) if the mean expression of the marker gene is bigger than the threshold of mean expression (thresh_mean) and if the fraction of cells with the marker gene expression is equal or higher than the cell fraction threshold (thresh_frac).
+
 More details on the `scanpy documentation
-<https://icb-scanpy.readthedocs-hosted.com/en/@version@/api/scanpy.pp.downsample_counts.html>`__
+<https://scanpy.readthedocs.io/en/stable/api/scanpy.pp.downsample_counts.html>`__
 
 
     ]]></help>