diff --git a/docs/tutorials/index.rst b/docs/tutorials/index.rst
index 6c81c40..9d59a94 100644
--- a/docs/tutorials/index.rst
+++ b/docs/tutorials/index.rst
@@ -15,6 +15,35 @@ Run scar with Python API
 
 Run scar with the command line tool
 ---------------------------------
+
+Use ``.h5`` files as the input 
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+We can directly use cellranger outputed *filtered_feature_bc_matrix.h5* to run ``scar`` by::
+
+   scar filtered_feature_bc_matrix.h5 -ft feature_type -o output
+
+``filtered_feature_bc_matrix.h5``, the cellranger output.
+
+``feature_type``, a string, either 'mRNA' or 'sgRNA' or 'ADT' or 'tag' or 'CMO'.
+
+ .. note::
+      The ambient profile is calculated by averaging the cell pool under this mode.
+      
+The output folder contains an h5ad file::
+   
+   output
+	└── filtered_feature_bc_matrix_denoised_feature_type.h5ad
+
+In the h5ad object:
+
+- anndata.X, denosed counts.  
+- anndata.obs[``noise_ratio``], estimated noise ratio per cell.  
+- anndata.layer[``native_frequencies``], estimated native frequencies.  
+- anndata.layer[``BayesFactor``], bayesian factor of ambient contamination.
+- anndata.obs[``sgRNAs``], optional, feature assignment, e.g., sgRNA, tag, CMO, and etc..
+
+Use ``.pickle`` files as the input 
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 We can also run ``scar`` by::
    
    scar raw_count_matrix.pickle -ft feature_type -o output
@@ -80,7 +109,4 @@ In the folder structure above:
 - ``denoised_counts.pickle``, denoised count matrix.
 - ``BayesFactor.pickle``, bayesian factor of ambient contamination.
 - ``expected_native_freq.pickle``, estimated native frequencies.  
-- ``assignment.pickle``, feature assignment, e.g., sgRNA, tag, and etc..
-
-.. note::
-    Only ``.pickle`` format is supported at the moment.
\ No newline at end of file
+- ``assignment.pickle``, optional, feature assignment, e.g., sgRNA, tag, and etc..
\ No newline at end of file
diff --git a/scar/main/__main__.py b/scar/main/__main__.py
index bd3fac7..1a4a02c 100644
--- a/scar/main/__main__.py
+++ b/scar/main/__main__.py
@@ -85,7 +85,7 @@ def main():
             adata_fb = adata[:, adata.var["feature_types"] == features]
             count_matrix = adata_fb.to_df()
 
-        print(f"{len(features)} modalities to denoise: {features}")
+        print(f"modalities to denoise: {features}")
 
     else:
         raise Exception(file_extension + " files are not supported.")
@@ -191,7 +191,7 @@ def main():
     elif file_extension == ".h5":
 
         output_path_h5ad = os.path.join(
-            output_dir, "filtered_feature_bc_matrix_denoised_{feature_type}.h5ad"
+            output_dir, f"filtered_feature_bc_matrix_denoised_{feature_type}.h5ad"
         )
 
         denoised_adata = adata.copy()
@@ -247,14 +247,14 @@ def scar_parser():
         "count_matrix",
         type=str,
         nargs="+",
-        help="The file of observed count matrix, 2D array (cells x genes) or the path of a filtered_feature_bc_matrix.h5",
+        help="The file of raw count matrix, 2D array (cells x genes) or the path of a filtered_feature_bc_matrix.h5",
     )
     parser.add_argument(
         "-ap",
         "--ambient_profile",
         type=str,
         default=None,
-        help="The file of empty profile obtained from empty droplets, 1D array or the path of a raw_feature_bc_matrix.h5",
+        help="The file of empty profile obtained from empty droplets, 1D array",
     )
     parser.add_argument(
         "-ft",