diff --git a/scripts/sbol_to_dataset.py b/scripts/sbol_to_dataset.py
new file mode 100644
index 0000000..a19c5e3
--- /dev/null
+++ b/scripts/sbol_to_dataset.py
@@ -0,0 +1,44 @@
+from sbol2 import Document
+import pandas as pd
+import argparse
+
+
+def sbol_to_csv(input_file, output_file="dataset.csv"):
+    doc = Document()
+    doc.read(input_file)
+
+    data = []
+
+    for comp in doc.componentDefinitions:
+        seq = None
+
+        try:
+            if comp.sequences:
+                seq_id = comp.sequences[0]
+                sequence_obj = doc.sequences.get(seq_id)
+
+                if sequence_obj:
+                    seq = sequence_obj.elements
+        except Exception:
+            continue
+
+        if seq:
+            data.append({
+                "sequence": seq,
+                "label": "unknown"
+            })
+
+    df = pd.DataFrame(data)
+    df.to_csv(output_file, index=False)
+
+    print(f"Dataset saved to {output_file}")
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser(description="Convert SBOL to CSV dataset")
+    parser.add_argument("--input", required=True, help="Input SBOL file")
+    parser.add_argument("--output", default="dataset.csv", help="Output CSV file")
+
+    args = parser.parse_args()
+
+    sbol_to_csv(args.input, args.output)
\ No newline at end of file