fixed errors- functions for easy clustering

mmseqs2/app.py

@@ -1,8 +1,10 @@
 from fastapi import FastAPI, HTTPException, Request
+from starlette.responses import FileResponse
 import logging
 
 import subprocess
 import os
+import shutil
 
 app = FastAPI()
 logging.basicConfig(level=logging.INFO)
@@ -60,21 +62,6 @@ def validate_sequence(sequence: str) -> bool:
 
     print(f"FASTA file created: {filename}")
 
-def create_queryDB_from_seq(filename):
-
-    command = [
-        "mmseqs", "createdb",
-        filename,
-        filename.replace('fasta', '') + ".db"
-    ]
-
-    try:
-        subprocess.run(command, check=True)
-
-    except subprocess.CalledProcessError as e:
-        raise HTTPException(status_code=600, detail=str(e))
-
-
 @app.get("/")
 async def read_root():
     return {"message": "Welcome to the MMSeqs2 API!"}
@@ -91,60 +78,14 @@ def help():
     except subprocess.CalledProcessError as e:
         raise HTTPException(status_code=400, detail=f"Command failed {e.stderr}")
 
-
-# @app.post("/easycluster")
-# async def easycluster(request: Request):
-
-#     logging.basicConfig(level=logging.INFO)
-#     logger = logging.getLogger(__name__)
-
-#     data = await request.json()
-#     logger.info(f"Received request data: {data}")
-
-#     print(f" Received request to run mmseqs with data: {data}")
-
-#     query_filename =f"in.fasta"
-#     result_filename = f"out"
-
-#     # Clear or create result file 
-#     open(result_filename, 'w').close()
-
-#     # Create the fasta file from the query string
-#     create_fastas_file_from_seq(data['query'], query_filename)
-
-#     # Run the mmseqs2 search command
-#     # command = [
-#     #     "mmseqs", "easy-cluster", 
-#     #     query_filename, 
-#     #     result_filename,
-#     #     "--min-seq-id", request['min_seq_id'], 
-#     #     "-c", request['coverage'], 
-#     #     "--cov-mode", request['cov_mode'],
-#     #     "tmp"
-#     # ]
-#     command = ["mmseqs", "easy-cluster", query_filename, result_filename, "tmp"]
-
-#     logger.info(f"Running command: {' '.join(command)}")
-
-#     try:
-#         subprocess.run(command, check=True)
-#     except subprocess.CalledProcessError as e:
-#         logger.error(f"MMSeqs command failed: {e}")
-#         raise HTTPException(status_code=500, detail=str(e))
-
-#     with open(result_filename, 'r') as file:
-#         result = file.read()
-
-#     return result
-
 @app.post("/easycluster")
 async def easycluster(request: Request):
     data = await request.json()
     logger.info(f"Received request data: {data}")
 
     BASE_DIR = "/app"
     query_filename = os.path.join(BASE_DIR, "in.fasta")
-    result_filename = os.path.join(BASE_DIR, "output.out")
+    result_filename = os.path.join(BASE_DIR, "output")
     tmp_dir = os.path.join(BASE_DIR, "tmp")
 
     os.makedirs(tmp_dir, exist_ok=True)
@@ -154,29 +95,32 @@ async def easycluster(request: Request):
     create_fastas_file_from_seq(data['query'], query_filename)
 
     # Run the mmseqs2 command
-    command = ["mmseqs", "easy-cluster", query_filename, result_filename, tmp_dir]
+    command = [
+        "mmseqs", 
+        "easy-cluster", 
+        query_filename, 
+        result_filename, 
+        '--min-seq-id', str(data['min_seq_id']),
+        '-c', str(data['coverage']),
+        '--cov-mode', str(data['cov_mode']),
+        tmp_dir]
     logger.info(f"Running command: {' '.join(command)}")
 
     try:
-        results = subprocess.run(command, capture_output=True, text=True, check=True)
-        logger.info(f"Command output: {results.stdout}")
+        result = subprocess.run(command, capture_output=True, text=True, check=True)
+        logger.info(f"Command output: {result.stdout}")
 
     except subprocess.CalledProcessError as e:
         logger.error(f"Command failed with return code {e.returncode}")
         logger.error(f"STDOUT: {e.stdout}")
         logger.error(f"STDERR: {e.stderr}")
         raise HTTPException(status_code=500, detail=f"Command failed: {e.stderr}")
-
-    # with open(f"/app/{result_filename}.out_all_seqs.fasta", 'r') as file:
-    #     result = file.read()
-
-    with open(result_filename, 'r') as file:
-        print(f"Reading result file: {result_filename}")
-        result = file.read()
-
-    return {"result": result}
 
-# add easy search
+    with open("/app/output_all_seqs.fasta", 'r') as file:
+        logger.info(f"Reading result file: /app/output_all_seqs.fasta")
+        result = file.read()
+
+    return result
 
 if __name__ == '__main__':
     import uvicorn