Merge branch 'master' of /home/git/public/genecup

1 files changed, 88 insertions, 16 deletions

diff --git a/server.py b/server.py
index c81cbc9..25ccad5 100755
--- a/server.py
+++ b/server.py
@@ -64,7 +64,7 @@ import re
 import ast
 from more_functions import *
 from nltk.tokenize import sent_tokenize
-from more_functions import getabstracts, undic, gene_category
+from more_functions import getabstracts, getabstracts_batch, undic, gene_category
 
 GENECUP_PROMPT_TEMPLATE = ""
 try:
@@ -346,6 +346,67 @@ def logout():
 def about():
     return render_template('about.html',version=version())
 
+@app.route("/create-ontology", methods=["GET", "POST"])
+def create_ontology():
+    if request.method == "GET":
+        session.pop('namecat', None)
+    from more_functions import gemini_query
+    default_prompt = (
+        "Give me a list of terms on substance abuse disorder (SUD) that act "
+        "as traits and classifiers in scientific literature with a focus on "
+        "behaviour and brain attributes related to the hippocampus. Avoid "
+        "aliases and synonyms as well as gene names. Each term should be "
+        "1-3 words (max). Give me a list of at least 20, but no more than "
+        "80, most used terms. Return only the terms, one per line, no "
+        "numbering. Add abbreviations and aliases - each at least 3 letters that have no other meaning - as a list with each term, "
+        "separated by commas")
+    if request.method == "POST":
+        action = request.form.get("action", "generate")
+
+        if action == "search":
+            # Build a temporary .onto file from the result terms and redirect to /progress
+            result_text = request.form.get("result", "")
+            query = request.form.get("query", "")
+            search_types = request.form.getlist("type")
+            # Build onto dict: each term is its own category with aliases as pipe-separated keywords
+            dict_onto = {}
+            for line in result_text.strip().split("\n"):
+                line = line.strip()
+                if not line:
+                    continue
+                parts = [p.strip() for p in line.split(",")]
+                category = parts[0]
+                keywords = "|".join(parts)
+                dict_onto[category] = {category: {keywords}}
+            # Save to a temp .onto file
+            onto_path = os.path.join(tempfile.gettempdir(), "gemini_ontology")
+            with open(onto_path + ".onto", "w") as f:
+                f.write(repr(dict_onto))
+            session['namecat'] = onto_path
+            print(f"  Created ontology: {onto_path}.onto with {len(dict_onto)} categories")
+            print(f"  Gene query: '{query}', search_types: {search_types}")
+            # Build the redirect URL with type and query params
+            from urllib.parse import urlencode
+            params = [("query", query)]
+            for t in search_types:
+                params.append(("type", t))
+            return redirect("/progress?" + urlencode(params))
+
+        # action == "generate"
+        prompt = request.form.get("prompt", default_prompt)
+        try:
+            result = gemini_query(prompt)
+            terms = [t.strip() for t in result.strip().split("\n") if t.strip()]
+            return render_template('create-ontology.html',
+                                   prompt=prompt, result=result,
+                                   count=len(terms), version=version())
+        except Exception as e:
+            return render_template('create-ontology.html',
+                                   prompt=prompt, result=f"Error: {e}",
+                                   count=0, version=version())
+    return render_template('create-ontology.html',
+                           prompt=default_prompt, result=None,
+                           count=0, version=version())
 
 # Ontology selection
 @app.route("/index_ontology", methods=["POST", "GET"])
@@ -803,6 +864,11 @@ def progress():
         if (search_type == []):
             search_type = ['GWAS', 'function', 'addiction', 'drug', 'brain', 'stress', 'psychiatric', 'cell']
         session['search_type'] = search_type
+        # Use default addiction ontology unless redirected from /create-ontology
+        if request.referrer and '/create-ontology' in request.referrer:
+            pass  # keep session['namecat'] set by /create-ontology
+        elif 'namecat' in session:
+            del session['namecat']
     genes_session = ''
 
     for gen in genes:
@@ -846,8 +912,10 @@ def search():
     if 'namecat' in session:
         namecat_flag=1
         ses_namecat = session['namecat']
+        print(f"  /search: namecat={ses_namecat}, search_type={search_type}")
         onto_cont = open(session['namecat']+".onto","r").read()
         dict_onto=ast.literal_eval(onto_cont)
+        print(f"  /search: onto categories={list(dict_onto.keys())[:10]}")
 
         for ky in dict_onto.keys():
             nodecolor[ky] = "hsl("+str((n_num+1)*int(360/len(dict_onto.keys())))+", 70%, 80%)"
@@ -908,28 +976,32 @@ def search():
                     all_d = all_d+'|'+all_d_ls
             if all_d: # Check if all_d is not empty
                 all_d=all_d[1:]
+            print(f"  /search generate: all_d={all_d[:200] if all_d else '(empty)'}, search_type={search_type}")
 
             if ("GWAS" in search_type):
                 datf = pd.read_csv('./utility/gwas_used.csv',sep='\t')
             progress+=percent
             yield "data:"+str(progress)+"\n\n"
 
-            for gene in genes:
-                print(f"Fetching info for gene {gene}\n")
-                abstracts_raw = getabstracts(gene,all_d) # all_d might be empty if no search_type matches
-                print(abstracts_raw)
-                sentences_ls=[]
+            # Batch fetch all abstracts in a single PubMed query
+            print(f"Batch fetching abstracts for {len(genes)} genes")
+            all_abstracts_raw = getabstracts_batch(genes, all_d) if all_d else ""
+            # Parse all sentences once
+            all_sentences = []
+            for row in all_abstracts_raw.split("\n"):
+                if not row.strip(): continue
+                tiab = row.split("\t")
+                pmid = tiab.pop(0)
+                tiab_text = " ".join(tiab)
+                for sent_tok in sent_tokenize(tiab_text):
+                    all_sentences.append(pmid + ' ' + sent_tok)
 
-                for row in abstracts_raw.split("\n"):
-                    if not row.strip(): continue # Skip empty lines
-                    tiab=row.split("\t")
-                    pmid = tiab.pop(0)
-                    tiab_text = " ".join(tiab) # Renamed to avoid conflict
-                    sentences_tok = sent_tokenize(tiab_text)
-                    for sent_tok in sentences_tok:
-                        sent_tok = pmid + ' ' + sent_tok
-                        sentences_ls.append(sent_tok)
-                gene=gene.replace("-"," ")
+            for gene in genes:
+                gene = gene.replace("-", " ")
+                # Filter sentences that mention this gene
+                gene_re = re.compile(r'\b' + re.escape(gene) + r'\b', re.IGNORECASE)
+                sentences_ls = [s for s in all_sentences if gene_re.search(s)]
+                print(f"  Gene {gene}: {len(sentences_ls)} sentences")
 
                 geneEdges = ""