From 82b9f84f71f5e8e83a99f90e8367127cc02c938f Mon Sep 17 00:00:00 2001
From: sepro <sebastian.proost@gmail.com>
Date: Wed, 29 Nov 2017 16:27:56 +0100
Subject: [PATCH] groundwork comparative heatmap

---
 README.md                                     |  4 +-
 conekt/controllers/heatmap.py                 | 20 +++++++++
 .../expression/cross_species_profile.py       | 42 ++++++++++++++++---
 3 files changed, 58 insertions(+), 8 deletions(-)

diff --git a/README.md b/README.md
index 70d432c..c42d259 100644
--- a/README.md
+++ b/README.md
@@ -11,8 +11,8 @@ networks and more across different species. A public version including seven pla
 publicly available RNASeq samples (from the [SRA](https://www.ncbi.nlm.nih.gov/sra)) is available 
 [here](http://conekt.mpimp-golm.mpg.de/pub/).
 
-Here you can find **Tutorials** how to use the public version along with the source-code and build instructions that 
-allow you to host your own instance with other species or in-house data.
+Here you can find **tutorials** how to use the public version along with the **source-code** and build instructions that 
+allow you to **host your own instance** with other species or in-house data.
 
 Tutorials
 ---------
diff --git a/conekt/controllers/heatmap.py b/conekt/controllers/heatmap.py
index 989fd15..dd19562 100644
--- a/conekt/controllers/heatmap.py
+++ b/conekt/controllers/heatmap.py
@@ -8,6 +8,8 @@
 from conekt.models.expression.profiles import ExpressionProfile
 from conekt.models.relationships.sequence_cluster import SequenceCoexpressionClusterAssociation
 from conekt.models.sequences import Sequence
+from conekt.models.trees import Tree
+from conekt.models.expression.cross_species_profile import CrossSpeciesExpressionProfile
 
 heatmap = Blueprint('heatmap', __name__)
 
@@ -93,6 +95,24 @@ def heatmap_main():
         return render_template("expression_heatmap.html", form=form, example=example)
 
 
+@heatmap.route('/comparative/tree/<int:tree_id>')
+@heatmap.route('/comparative/tree/<int:tree_id>/<option>')
+@cache.cached()
+def heatmap_comparative_tree(tree_id, option='raw'):
+    tree = Tree.query.get_or_404(tree_id)
+    sequences = tree.sequences
+    sequence_ids = [s.id for s in sequences]
+
+    heatmap_data = CrossSpeciesExpressionProfile().get_heatmap(*sequence_ids, option=option)
+
+    print(heatmap_data)
+
+    return render_template("expression_heatmap.html", order=heatmap_data['order'],
+                           profiles=heatmap_data['heatmap_data'],
+                           zlog=1 if option == 'zlog' else 0,
+                           raw=1 if option == 'raw' else 0)
+
+
 @heatmap.route('/inchlib/j/<cluster_id>.json')
 @cache.cached()
 def heatmap_inchlib_json(cluster_id):
diff --git a/conekt/models/expression/cross_species_profile.py b/conekt/models/expression/cross_species_profile.py
index c0cb320..bd430d6 100644
--- a/conekt/models/expression/cross_species_profile.py
+++ b/conekt/models/expression/cross_species_profile.py
@@ -38,26 +38,31 @@ def get_data(self, *sequence_ids):
 
                 parsed_profile = {
                     "order": self.conditions,
-                    "data": {c: max(current_profile["data"][c]) if c in current_profile["data"].keys() else None
+                    "data": {},
+                    "raw_data": {c: max(current_profile["data"][c]) if c in current_profile["data"].keys() else None
                              for c in self.conditions}
                     }
 
                 # detect low expressed genes before normalization
                 low_expressed = all(
-                    [value < 10 for value in parsed_profile["data"].values() if value is not None])
+                    [value < 10 for value in parsed_profile["raw_data"].values() if value is not None])
 
                 # normalize profile
-                min_value = min([i if i is not None else 0 for i in parsed_profile["data"].values()])
-                max_value = max([i if i is not None else 0 for i in parsed_profile["data"].values()])
+                min_value = min([i if i is not None else 0 for i in parsed_profile["raw_data"].values()])
+                max_value = max([i if i is not None else 0 for i in parsed_profile["raw_data"].values()])
 
                 if max_value > 0:
                     for c in self.conditions:
-                        if parsed_profile["data"][c] is not None:
-                            parsed_profile["data"][c] = parsed_profile["data"][c]/max_value
+                        if parsed_profile["raw_data"][c] is not None:
+                            parsed_profile["data"][c] = parsed_profile["raw_data"][c]/max_value
+                        else:
+                            parsed_profile["data"][c] = None
 
                 converted_profiles.append(
                     {
                         "sequence_id": p.sequence_id,
+                        "sequence_name": p.sequence.name if p.sequence is not None else None,
+                        "shortest_alias": p.sequence.shortest_alias if p.sequence is not None else None,
                         "species_id": p.species_id,
                         "low_expressed": 1 if low_expressed else 0,
                         "profile": parsed_profile,
@@ -67,3 +72,28 @@ def get_data(self, *sequence_ids):
                 )
 
         return converted_profiles
+
+    def get_heatmap(self, *sequence_ids, option='raw'):
+        data = self.get_data(*sequence_ids)
+
+        output = {
+            'order': [],
+            'heatmap_data': []
+        }
+
+        for d in data:
+            if "profile" in d.keys() and "order" in d["profile"].keys():
+                output['order'] = d["profile"]["order"]
+                break
+
+        key = 'raw_data' if option == 'raw' else 'data'
+
+        for d in data:
+            output['heatmap_data'].append({
+                'sequence_id': d['sequence_id'],
+                'name': d['sequence_name'],
+                'shortest_alias': d['shortest_alias'],
+                'values': {k: v if v is not None else '-' for k,v in d['profile'][key].items()}
+            })
+
+        return output