diff --git a/gears/data_utils.py b/gears/data_utils.py
index f97972f..463b917 100644
--- a/gears/data_utils.py
+++ b/gears/data_utils.py
@@ -81,7 +81,12 @@ def get_dropout_non_zero_genes(adata):
     for i, j in conditions2index.items():
         condition2mean_expression[i] = np.mean(adata.X[j], axis = 0)
     pert_list = np.array(list(condition2mean_expression.keys()))
-    mean_expression = np.array(list(condition2mean_expression.values())).reshape(len(adata.obs.condition.unique()), adata.X.toarray().shape[1])
+    # to handle the non-sparse data input
+    try :
+        adata.X = adata.X.toarray()
+        mean_expression = np.array(list(condition2mean_expression.values())).reshape(len(adata.obs.condition.unique()), adata.X.shape[1])
+    except:
+        mean_expression = np.array(list(condition2mean_expression.values())).reshape(len(adata.obs.condition.unique()), adata.X.shape[1])
     ctrl = mean_expression[np.where(pert_list == 'ctrl')[0]]
     
     ## in silico modeling and upperbounding
@@ -408,4 +413,4 @@ def get_genes_from_perts(self, perts):
         gene_list = [p.split('+') for p in np.unique(perts)]
         gene_list = [item for sublist in gene_list for item in sublist]
         gene_list = [g for g in gene_list if g != 'ctrl']
-        return np.unique(gene_list)
\ No newline at end of file
+        return np.unique(gene_list)
diff --git a/gears/pertdata.py b/gears/pertdata.py
index 801f289..232387c 100644
--- a/gears/pertdata.py
+++ b/gears/pertdata.py
@@ -594,8 +594,13 @@ def create_cell_graph_dataset(self, split_adata, pert_category,
         # Create cell graphs
         cell_graphs = []
         for X, y in zip(Xs, ys):
-            cell_graphs.append(self.create_cell_graph(X.toarray(),
-                                y.toarray(), de_idx, pert_category, pert_idx))
+            try:
+                cell_graphs.append(self.create_cell_graph(X.toarray(),
+                                    y, de_idx, pert_category, pert_idx))
+            except:
+                y = y.toarray()
+                cell_graphs.append(self.create_cell_graph(X.toarray(),
+                                    y, de_idx, pert_category, pert_idx))
 
         return cell_graphs