diff --git a/Ab_epitope/.gitignore b/Ab_epitope/.gitignore
new file mode 100644
index 0000000..5d81013
--- /dev/null
+++ b/Ab_epitope/.gitignore
@@ -0,0 +1,9 @@
+OAS_download
+
+data/epitope_all.fasta
+
+result/
+
+data/dataset
+
+data/raw_data/OAS_memory_paired.csv
diff --git a/Ab_epitope/data_clean_for_LM.sh b/Ab_epitope/data_clean_for_LM.sh
index 52b756e..01220b5 100755
--- a/Ab_epitope/data_clean_for_LM.sh
+++ b/Ab_epitope/data_clean_for_LM.sh
@@ -1,25 +1,40 @@
 #!/bin/bash
-python mBLM/script/merge_raw_data.py
-python mBLM/script/df2fasta.py mBLM/result/memory_paired_Abs.csv mBLM/result/memory_paired_Abs.fasta
-
 # clustering
 output_prefix=mBLM/result/cluster/
 f=mBLM/result/memory_paired_Abs.fasta
 name=memory_paired_Abs
+
+echo "Checking if $output_prefix exists"
+if [ ! -d "$output_prefix" ]; then
+	echo "Creating output directory"
+	mkdir -p "$output_prefix"
+fi
+
+python mBLM/script/merge_raw_data.py
+python mBLM/script/df2fasta.py mBLM/result/memory_paired_Abs.csv mBLM/result/memory_paired_Abs.fasta
+
+
 for x in 0.5 0.6
 do
-	cd-hit -i $f -o $output_prefix$name$x -c $x -M 32000 -d 0 -T 32 -n 3 -aL 0.8 -s 0.95  -uS 0.2  -sc 1 -sf 1
+	cd-hit -i $f -o $output_prefix$name$x -c $x -M 32000 -d 0 -T 32 -n 3 -aL 0.8 -s 0.95 -uS 0.2 -sc 1 -sf 1
 done
     
 for x in 0.7 0.8 0.9 0.95
 do
-	cd-hit -i $f -o $output_prefix$name$x -c $x -M 32000 -d 0 -T 32 -n 5 -aL 0.8 -s 0.95  -uS 0.2  -sc 1 -sf 1
+	cd-hit -i $f -o $output_prefix$name$x -c $x -M 32000 -d 0 -T 32 -n 5 -aL 0.8 -s 0.95 -uS 0.2 -sc 1 -sf 1
 done  
 
 
 # add cluster id to dataset
-
 python mBLM/script/add_clstr2df.py -i mBLM/result/memory_paired_Abs.csv -o mBLM/result/memory_paired_Abs_final.csv
 
+
+datasetdir=data/dataset 
+echo "Checking if $datasetdir exists"
+if [ ! -d "$datasetdir" ]; then
+	echo "Creating output directory"
+	mkdir "$datasetdir"
+fi
+
 # split dataset
 python mBLM/script/split_dataset.py -i mBLM/result/memory_paired_Abs_final.csv -o data/dataset/memory_paired_Abs
diff --git a/Ab_epitope/data_clean_for_epitope.sh b/Ab_epitope/data_clean_for_epitope.sh
index 2b55088..1d7904b 100644
--- a/Ab_epitope/data_clean_for_epitope.sh
+++ b/Ab_epitope/data_clean_for_epitope.sh
@@ -10,9 +10,16 @@ fasta_prefix=./data/
 output_prefix=./result/cluster/
 f=epitope_all.fasta
 name=`basename $f ".fasta"`
+
+echo "Checking if $output_prefix exists"
+if [ ! -d "$output_prefix" ]; then
+	echo "Creating output directory"
+	mkdir "$output_prefix"
+fi
+
 for x in 0.8 0.9 0.92 0.95 0.98 0.99 0.995
 do
-	cd-hit -i $f -o $output_prefix$name$x$fasta_suffix -c $x -M 32000 -d 0 -T 8 -n 5 -aL 0.8 -s 0.95  -uS 0.2  -sc 1 -sf 1
+	cd-hit -i $fasta_prefix$f -o $output_prefix$name$x$fasta_suffix -c $x -M 32000 -d 0 -T 8 -n 5 -aL 0.8 -s 0.95  -uS 0.2  -sc 1 -sf 1
 done
 
 
@@ -23,4 +30,4 @@ python ./script/add_clstr2df.py
 # split dataset
 python ./script/split_dataset.py 
 # remove repeat fasta sequence
-pytho script/rm_fas_repeats.py result/epitope_clean.fasta result/epitope_clean_v2.fasta 
\ No newline at end of file
+python script/rm_fas_repeats.py result/epitope_clean.fasta result/epitope_clean_v2.fasta 
\ No newline at end of file