Merge pull request #81 from daisybio/dev

combined sorting

Author: strasserle

Diff for: app/controllers/geneInteraction.py

@@ -1,7 +1,7 @@
 import sqlalchemy as sa
 import os
 from flask import jsonify
-from sqlalchemy import desc, or_, and_
+from sqlalchemy import desc, literal_column, or_, and_
 from sqlalchemy.sql import text
 from app.controllers.dataset import _dataset_query
 import app.models as models
@@ -1017,7 +1017,7 @@ def getGeneCounts(dataset_ID: int = None, disease_name=None, ensg_number=None, g
 def get_gene_network(dataset_ID: int = None, disease_name=None,
                       minBetweenness:float = None, minNodeDegree:float = None, minEigenvector:float = None,
                       maxPValue=0.05, minMscor=None, minCorrelation=None,
-                      edgeSorting: str = None, nodeSorting: str = None,
+                      edgeSorting: str = None, nodeSorting: list[str] = None,
                       maxNodes: int = 100, maxEdges: int = 100, 
                       offsetNodes: int = None, offsetEdges: int = None, 
                       sponge_db_version: int = LATEST):
@@ -1032,7 +1032,7 @@ def get_gene_network(dataset_ID: int = None, disease_name=None,
     :param minMscor: mscor cutoff (>)
     :param minCorrelation: correlation cutoff (>)
     :param edgeSorting: sorting key for edges
-    :param nodeSorting: sorting key for nodes
+    :param nodeSorting: sorting key(s) for nodes (options are 'betweenness', 'degree', 'eigenvector')
     :param maxNodes: maximum number of nodes
     :param maxEdges: maximum number of edges
     :param offsetNodes: offset for node pagination
@@ -1080,17 +1080,15 @@ def get_gene_network(dataset_ID: int = None, disease_name=None,
     if minEigenvector:
         node_query = node_query.filter(models.networkAnalysis.eigenvector >= minEigenvector)
 
-    # Sorting nodes
-    if nodeSorting == "betweenness":
-        node_query = node_query.order_by(models.networkAnalysis.betweenness.desc())
-    elif nodeSorting == "degree":
-        node_query = node_query.order_by(models.networkAnalysis.node_degree.desc())
-    elif nodeSorting == "eigenvector":
-        node_query = node_query.order_by(models.networkAnalysis.eigenvector.desc())
-    else:
-        raise ValueError("Invalid node sorting key. Choose one of 'betweenness', 'degree', 'eigenvector'")
+    # Sorting nodes: if more than one sorting key, rank by each key individually and sort by the mean of the ranks
+    if nodeSorting:
+        if any([key not in ['betweenness', 'node_degree', 'eigenvector'] for key in nodeSorting]):
+            raise ValueError("Invalid node sorting key. Choose from 'betweenness', 'node_degree', 'eigenvector'")
+        rank_columns = [db.func.rank().over(order_by=getattr(models.networkAnalysis, col).desc()).label(f"{col}_rank") for col in nodeSorting]
+        mean_rank = sum(rank_columns) / len(rank_columns)
+        node_query = node_query.order_by(mean_rank)
 
-    # node agination
+    # node pagination
     node_query = node_query.offset(offsetNodes).limit(maxNodes)
 
     # filter edges based on filtered nodes

Diff for: app/controllers/transcriptInteraction.py

@@ -845,7 +845,7 @@ def getTranscriptCounts(dataset_ID: int = None, disease_name=None, enst_number=N
 def get_transcript_network(dataset_ID: int = None, disease_name=None,
                             minBetweenness:float = None, minNodeDegree:float = None, minEigenvector:float = None,
                             maxPValue=0.05, minMscor=None, minCorrelation=None,
-                            edgeSorting: str = None, nodeSorting: str = None,
+                            edgeSorting: str = None, nodeSorting: list[str] = None,
                             maxNodes: int = 100, maxEdges: int = 100, 
                             offsetNodes: int = None, offsetEdges: int = None,
                             sponge_db_version: int = LATEST):
@@ -860,7 +860,7 @@ def get_transcript_network(dataset_ID: int = None, disease_name=None,
     :param minMscor: mscor cutoff (>)
     :param minCorrelation: correlation cutoff (>)
     :param edgeSorting: sorting key for edges
-    :param nodeSorting: sorting key for nodes
+    :param nodeSorting: sorting key(s) for nodes (options are 'betweenness', 'node_degree', 'eigenvector')
     :param maxNodes: maximum number of nodes
     :param maxEdges: maximum number of edges
     :param offsetNodes: offset for node pagination
@@ -910,15 +910,13 @@ def get_transcript_network(dataset_ID: int = None, disease_name=None,
     if minEigenvector:
         node_query = node_query.filter(models.networkAnalysisTranscript.eigenvector >= minEigenvector)
 
-    # Sorting nodes 
-    if nodeSorting == "betweenness":
-        node_query = node_query.order_by(models.networkAnalysisTranscript.betweenness.desc())
-    elif nodeSorting == "degree":
-        node_query = node_query.order_by(models.networkAnalysisTranscript.node_degree.desc())
-    elif nodeSorting == "eigenvector":
-        node_query = node_query.order_by(models.networkAnalysisTranscript.eigenvector.desc())
-    else: 
-        raise ValueError("Invalid node sorting key. Choose one of 'betweenness', 'degree', 'eigenvector'")
+    # Sorting nodes: if more than one sorting key, rank by each key individually and sort by the mean of the ranks
+    if nodeSorting:
+        if any([key not in ['betweenness', 'node_degree', 'eigenvector'] for key in nodeSorting]):
+            raise ValueError("Invalid node sorting key. Choose from 'betweenness', 'node_degree', 'eigenvector'")
+        rank_columns = [db.func.rank().over(order_by=getattr(models.networkAnalysisTranscript, col).desc()).label(f"{col}_rank") for col in nodeSorting]
+        mean_rank = sum(rank_columns) / len(rank_columns)
+        node_query = node_query.order_by(mean_rank)
 
     # node pagination
     node_query = node_query.offset(offsetNodes).limit(maxNodes)    

Diff for: swagger.yml

@@ -1432,11 +1432,13 @@ paths:
           required: false
           description: Possibilities for sorting of the nodes.
           schema:
-            type: string
-            enum:
-              - betweenness
-              - degree
-              - eigenvector
+            type: array
+            items:
+              type: string
+            # enum:
+            #   - betweenness
+            #   - degree
+            #   - eigenvector
         - name: maxNodes
           in: query
           description: Number of nodes that should be shown. Default value is 100 and can be up to 1000. For more results please use batches, the provided offset parameter or download the whole dataset.
@@ -1552,11 +1554,13 @@ paths:
           required: false
           description: Possibilities for sorting of the nodes.
           schema:
-            type: string
-            enum:
-              - betweenness
-              - degree
-              - eigenvector
+            type: array
+            items:
+              type: string
+            # enum:
+            #   - betweenness
+            #   - degree
+            #   - eigenvector
         - name: maxNodes
           in: query
           description: Number of nodes that should be shown. Default value is 100 and can be up to 1000. For more results please use batches, the provided offset parameter or download the whole dataset.
@@ -1714,7 +1718,7 @@ paths:
           type: string
           enum:
           - betweenness
-          - degree
+          - node_degree
           - eigenvector
       - name: descending
         in: query
@@ -1895,7 +1899,7 @@ paths:
             type: string
             enum:
               - betweenness
-              - degree
+              - node_degree
               - eigenvector
         - name: descending
           in: query