fuse.py: add a limit to similarity (don't compute gigantic similarity matrix)

tools/fuse.py

@@ -46,6 +46,7 @@ from collections import defaultdict
 FUSE_VERSION = "vidjil fuse"
 
 TOOL_SIMILARITY = "../algo/tools/similarity"
+SIMILARITY_LIMIT = 1000
 
 GERMLINES_ORDER = ['TRA', 'TRB', 'TRG', 'TRD', 'DD', 'IGH', 'DHJH', 'IJK', 'IJL'] 
 
@@ -739,6 +740,7 @@ def main():
     print()
 
     #compute similarity matrix
+    if len(jlist_fused.d["clones"]) < SIMILARITY_LIMIT :
         fasta = ""
         for i in range(len(jlist_fused.d["clones"])) :
             fasta += ">>" + str(i) + "\n"
@@ -750,6 +752,8 @@ def main():
             jlist_fused.d["similarity"] = json.loads(out)
         except OSError:
             print("! failed: %s" % TOOL_SIMILARITY)
+        
+        
     print("### Save merged file")
     jlist_fused.save_json(args.output)
     os.unlink(fasta_file.name)