diff --git a/srcs/hypothesis_recovery_src.py b/srcs/hypothesis_recovery_src.py
index a542b5e..483c24e 100644
--- a/srcs/hypothesis_recovery_src.py
+++ b/srcs/hypothesis_recovery_src.py
@@ -254,8 +254,11 @@ def hypothesis_recovery(
     sample_dir = os.path.dirname(sample_file)
     sample_name = os.path.basename(sample_file).replace('.sig.zip', '')
     path_to_sample_temp_dir = os.path.join(sample_dir, f'sample_{sample_name}_intermediate_files')
-    if not os.path.exists(path_to_sample_temp_dir):
-        os.makedirs(path_to_sample_temp_dir)
+    if os.path.exists(path_to_sample_temp_dir):
+        # if exists, remove it
+        logger.info(f"Removing existing temporary directory: {path_to_sample_temp_dir}")
+        os.system(f'rm -rf {path_to_sample_temp_dir}')
+    os.makedirs(path_to_sample_temp_dir)
     
     # Find the organisms that have non-zero overlap with the sample
     nontrivial_organism_names = get_organisms_with_nonzero_overlap(manifest, sample_file, scale, ksize, num_threads, path_to_genome_temp_dir, path_to_sample_temp_dir)
diff --git a/srcs/utils.py b/srcs/utils.py
index 4b7d974..34808f4 100644
--- a/srcs/utils.py
+++ b/srcs/utils.py
@@ -113,6 +113,10 @@ def run_multisearch(num_threads: int, ani_thresh: float, ksize: int, scale: int,
     B_TO_A = A_TO_B[['match_name','query_name']].rename(columns={'match_name':'query_name','query_name':'match_name'})
     multisearch_result = pd.concat([A_TO_B, B_TO_A]).drop_duplicates().reset_index(drop=True)
     
+    # change column type to string
+    multisearch_result['query_name'] = multisearch_result['query_name'].astype(str)
+    multisearch_result['match_name'] = multisearch_result['match_name'].astype(str)
+    
     return multisearch_result
 
 def remove_corr_organisms_from_ref(sig_info_dict: Dict[str, Tuple[str, float, int, int]], multisearch_result: pd.DataFrame) -> Tuple[Dict[str, List[str]], pd.DataFrame]: