Read IDs in the JSON file which saves the single molecule prediction output are truncated by 4 digits compared to the full read IDs.