Sepideh2027 commited on
Commit
19a7d92
·
verified ·
1 Parent(s): a2d72c7

Create app.py

Browse files
Files changed (1) hide show
  1. app.py +152 -0
app.py ADDED
@@ -0,0 +1,152 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import gradio as gr
2
+ import requests
3
+ import time
4
+ import re
5
+ from Bio import Entrez, SeqIO
6
+ from io import StringIO
7
+
8
+ Entrez.email = "anonymous@pathogenagent.ai"
9
+ Entrez.tool = "PathogenAgent-Lite"
10
+
11
+ HEADERS = {"User-Agent": "PathogenAgent-Lite/1.0"}
12
+
13
+ def search_pubmed(query, retmax=3):
14
+ try:
15
+ time.sleep(0.3)
16
+ url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi"
17
+ params = {"db": "pubmed", "term": query, "retmax": retmax, "retmode": "json"}
18
+ resp = requests.get(url, params=params, headers=HEADERS, timeout=10)
19
+ ids = resp.json().get("esearchresult", {}).get("idlist", [])
20
+ if not ids:
21
+ return []
22
+ url2 = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi"
23
+ params2 = {"db": "pubmed", "id": ",".join(ids), "retmode": "json"}
24
+ resp2 = requests.get(url2, params=params2, headers=HEADERS, timeout=10)
25
+ data = resp2.json()
26
+ results = []
27
+ for uid, rec in data.get("result", {}).items():
28
+ if uid == "uids":
29
+ continue
30
+ title = rec.get("title", "")
31
+ if title:
32
+ results.append(f"- {title}")
33
+ return results
34
+ except Exception:
35
+ return []
36
+
37
+ def search_genbank(query):
38
+ try:
39
+ time.sleep(0.5)
40
+ # بهبود جستجو: حذف کلمه "genome" و جستجوی نام موجودیت
41
+ search_term = query.replace("genome", "").strip()
42
+ if not search_term:
43
+ search_term = query
44
+
45
+ # جستجو با نام موجودیت
46
+ handle = Entrez.esearch(db="nucleotide", term=f"{search_term}[Organism]", retmax=3)
47
+ record = Entrez.read(handle)
48
+ handle.close()
49
+
50
+ if not record.get("IdList"):
51
+ return None
52
+
53
+ # گرفتن اولین نتیجه
54
+ seq_id = record["IdList"][0]
55
+ handle = Entrez.efetch(db="nucleotide", id=seq_id, rettype="fasta", retmode="text")
56
+ seq = SeqIO.read(StringIO(handle.read()), "fasta")
57
+ return {
58
+ "id": seq.id,
59
+ "length": len(seq.seq),
60
+ "snippet": str(seq.seq[:200])
61
+ }
62
+ except Exception:
63
+ return None
64
+
65
+ def search_clinvar(query):
66
+ try:
67
+ time.sleep(0.3)
68
+ url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi"
69
+ # برای ClinVar، فقط در صورتی که query شبیه واریانت باشد جستجو کن
70
+ if not re.search(r"c\.[0-9]+|rs[0-9]+|del|dup|ins|[A-Z][0-9]+[A-Z]", query):
71
+ return []
72
+ params = {"db": "clinvar", "term": query, "retmax": 3, "retmode": "json"}
73
+ resp = requests.get(url, params=params, headers=HEADERS, timeout=10)
74
+ ids = resp.json().get("esearchresult", {}).get("idlist", [])
75
+ if not ids:
76
+ return []
77
+ url2 = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi"
78
+ params2 = {"db": "clinvar", "id": ",".join(ids), "retmode": "json"}
79
+ resp2 = requests.get(url2, params=params2, headers=HEADERS, timeout=10)
80
+ data = resp2.json()
81
+ results = []
82
+ for uid, rec in data.get("result", {}).items():
83
+ if uid == "uids":
84
+ continue
85
+ title = rec.get("title", "")
86
+ if title:
87
+ results.append(f"- {title}")
88
+ return results
89
+ except Exception:
90
+ return []
91
+
92
+ def route_query(query):
93
+ q = query.lower()
94
+ if re.search(r"c\.[0-9]+|rs[0-9]+|del|dup|ins|[A-Z][0-9]+[A-Z]", q):
95
+ return "variant"
96
+ if re.search(r"genome|sequence|virus|organism|strain|cov|sars|ebola|influenza|mers|hiv|dengue|zika", q):
97
+ return "pathogen"
98
+ return "general"
99
+
100
+ def pathogen_agent(query):
101
+ route = route_query(query)
102
+ pubmed = search_pubmed(query)
103
+ clinvar = search_clinvar(query) if route == "variant" else []
104
+ genbank = search_genbank(query) if route == "pathogen" else None
105
+
106
+ pubmed_text = "\n".join(pubmed) if pubmed else "No PubMed results."
107
+ clinvar_text = "\n".join(clinvar) if clinvar else "No ClinVar results."
108
+
109
+ genbank_text = "Not applicable."
110
+ if genbank:
111
+ genbank_text = f"ID: {genbank['id']}\nLength: {genbank['length']:,} bp\n{genbank['snippet']}..."
112
+
113
+ return f"""
114
+ # 🧬 PathogenAgent (Lite - Hugging Face)
115
+
116
+ **Query:** {query}
117
+ **Type:** {route}
118
+
119
+ ---
120
+
121
+ ## 📄 PubMed
122
+ {pubmed_text}
123
+
124
+ ---
125
+
126
+ ## 🧪 ClinVar
127
+ {clinvar_text}
128
+
129
+ ---
130
+
131
+ ## 🧬 GenBank
132
+ {genbank_text}
133
+
134
+ ---
135
+
136
+ Generated by PathogenAgent
137
+ """
138
+
139
+ def run(query):
140
+ if not query.strip():
141
+ return "Please enter a query."
142
+ return pathogen_agent(query)
143
+
144
+ demo = gr.Interface(
145
+ fn=run,
146
+ inputs=gr.Textbox(label="Query", placeholder="e.g., SARS-CoV-2 genome or CFTR F508del"),
147
+ outputs=gr.Markdown(),
148
+ title="PathogenAgent (Hugging Face)",
149
+ description="Agentic AI for Pathogen Genomics"
150
+ )
151
+
152
+ demo.launch()