If you can't change the data directly, you must further carry out supplementary experiments in the process of revision, provide supplementary data and supporting information, provide reasonable and well-founded data, and don't make academic fraud because you avoid being consistent with the data of the big paper.